Training in progress, step 1300, checkpoint
Browse files- checkpoint-1300/README.md +202 -0
- checkpoint-1300/adapter_config.json +34 -0
- checkpoint-1300/adapter_model.safetensors +3 -0
- checkpoint-1300/global_step1299/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1300/global_step1299/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1300/global_step1299/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1300/global_step1299/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1300/global_step1299/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1300/global_step1299/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1300/global_step1299/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1300/global_step1299/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1300/global_step1299/mp_rank_00_model_states.pt +3 -0
- checkpoint-1300/latest +1 -0
- checkpoint-1300/rng_state_0.pth +3 -0
- checkpoint-1300/rng_state_1.pth +3 -0
- checkpoint-1300/rng_state_2.pth +3 -0
- checkpoint-1300/rng_state_3.pth +3 -0
- checkpoint-1300/rng_state_4.pth +3 -0
- checkpoint-1300/rng_state_5.pth +3 -0
- checkpoint-1300/rng_state_6.pth +3 -0
- checkpoint-1300/rng_state_7.pth +3 -0
- checkpoint-1300/scheduler.pt +3 -0
- checkpoint-1300/special_tokens_map.json +30 -0
- checkpoint-1300/tokenizer.json +0 -0
- checkpoint-1300/tokenizer_config.json +133 -0
- checkpoint-1300/trainer_state.json +2399 -0
- checkpoint-1300/training_args.bin +3 -0
- checkpoint-1300/zero_to_fp32.py +674 -0
checkpoint-1300/README.md
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: microsoft/Phi-3-mini-4k-instruct
|
| 3 |
+
library_name: peft
|
| 4 |
+
---
|
| 5 |
+
|
| 6 |
+
# Model Card for Model ID
|
| 7 |
+
|
| 8 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
| 9 |
+
|
| 10 |
+
|
| 11 |
+
|
| 12 |
+
## Model Details
|
| 13 |
+
|
| 14 |
+
### Model Description
|
| 15 |
+
|
| 16 |
+
<!-- Provide a longer summary of what this model is. -->
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- **Developed by:** [More Information Needed]
|
| 21 |
+
- **Funded by [optional]:** [More Information Needed]
|
| 22 |
+
- **Shared by [optional]:** [More Information Needed]
|
| 23 |
+
- **Model type:** [More Information Needed]
|
| 24 |
+
- **Language(s) (NLP):** [More Information Needed]
|
| 25 |
+
- **License:** [More Information Needed]
|
| 26 |
+
- **Finetuned from model [optional]:** [More Information Needed]
|
| 27 |
+
|
| 28 |
+
### Model Sources [optional]
|
| 29 |
+
|
| 30 |
+
<!-- Provide the basic links for the model. -->
|
| 31 |
+
|
| 32 |
+
- **Repository:** [More Information Needed]
|
| 33 |
+
- **Paper [optional]:** [More Information Needed]
|
| 34 |
+
- **Demo [optional]:** [More Information Needed]
|
| 35 |
+
|
| 36 |
+
## Uses
|
| 37 |
+
|
| 38 |
+
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
| 39 |
+
|
| 40 |
+
### Direct Use
|
| 41 |
+
|
| 42 |
+
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
|
| 43 |
+
|
| 44 |
+
[More Information Needed]
|
| 45 |
+
|
| 46 |
+
### Downstream Use [optional]
|
| 47 |
+
|
| 48 |
+
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
| 49 |
+
|
| 50 |
+
[More Information Needed]
|
| 51 |
+
|
| 52 |
+
### Out-of-Scope Use
|
| 53 |
+
|
| 54 |
+
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
| 55 |
+
|
| 56 |
+
[More Information Needed]
|
| 57 |
+
|
| 58 |
+
## Bias, Risks, and Limitations
|
| 59 |
+
|
| 60 |
+
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
| 61 |
+
|
| 62 |
+
[More Information Needed]
|
| 63 |
+
|
| 64 |
+
### Recommendations
|
| 65 |
+
|
| 66 |
+
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
|
| 67 |
+
|
| 68 |
+
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
|
| 69 |
+
|
| 70 |
+
## How to Get Started with the Model
|
| 71 |
+
|
| 72 |
+
Use the code below to get started with the model.
|
| 73 |
+
|
| 74 |
+
[More Information Needed]
|
| 75 |
+
|
| 76 |
+
## Training Details
|
| 77 |
+
|
| 78 |
+
### Training Data
|
| 79 |
+
|
| 80 |
+
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
| 81 |
+
|
| 82 |
+
[More Information Needed]
|
| 83 |
+
|
| 84 |
+
### Training Procedure
|
| 85 |
+
|
| 86 |
+
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
| 87 |
+
|
| 88 |
+
#### Preprocessing [optional]
|
| 89 |
+
|
| 90 |
+
[More Information Needed]
|
| 91 |
+
|
| 92 |
+
|
| 93 |
+
#### Training Hyperparameters
|
| 94 |
+
|
| 95 |
+
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
| 96 |
+
|
| 97 |
+
#### Speeds, Sizes, Times [optional]
|
| 98 |
+
|
| 99 |
+
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 100 |
+
|
| 101 |
+
[More Information Needed]
|
| 102 |
+
|
| 103 |
+
## Evaluation
|
| 104 |
+
|
| 105 |
+
<!-- This section describes the evaluation protocols and provides the results. -->
|
| 106 |
+
|
| 107 |
+
### Testing Data, Factors & Metrics
|
| 108 |
+
|
| 109 |
+
#### Testing Data
|
| 110 |
+
|
| 111 |
+
<!-- This should link to a Dataset Card if possible. -->
|
| 112 |
+
|
| 113 |
+
[More Information Needed]
|
| 114 |
+
|
| 115 |
+
#### Factors
|
| 116 |
+
|
| 117 |
+
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 118 |
+
|
| 119 |
+
[More Information Needed]
|
| 120 |
+
|
| 121 |
+
#### Metrics
|
| 122 |
+
|
| 123 |
+
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 124 |
+
|
| 125 |
+
[More Information Needed]
|
| 126 |
+
|
| 127 |
+
### Results
|
| 128 |
+
|
| 129 |
+
[More Information Needed]
|
| 130 |
+
|
| 131 |
+
#### Summary
|
| 132 |
+
|
| 133 |
+
|
| 134 |
+
|
| 135 |
+
## Model Examination [optional]
|
| 136 |
+
|
| 137 |
+
<!-- Relevant interpretability work for the model goes here -->
|
| 138 |
+
|
| 139 |
+
[More Information Needed]
|
| 140 |
+
|
| 141 |
+
## Environmental Impact
|
| 142 |
+
|
| 143 |
+
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 144 |
+
|
| 145 |
+
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
| 146 |
+
|
| 147 |
+
- **Hardware Type:** [More Information Needed]
|
| 148 |
+
- **Hours used:** [More Information Needed]
|
| 149 |
+
- **Cloud Provider:** [More Information Needed]
|
| 150 |
+
- **Compute Region:** [More Information Needed]
|
| 151 |
+
- **Carbon Emitted:** [More Information Needed]
|
| 152 |
+
|
| 153 |
+
## Technical Specifications [optional]
|
| 154 |
+
|
| 155 |
+
### Model Architecture and Objective
|
| 156 |
+
|
| 157 |
+
[More Information Needed]
|
| 158 |
+
|
| 159 |
+
### Compute Infrastructure
|
| 160 |
+
|
| 161 |
+
[More Information Needed]
|
| 162 |
+
|
| 163 |
+
#### Hardware
|
| 164 |
+
|
| 165 |
+
[More Information Needed]
|
| 166 |
+
|
| 167 |
+
#### Software
|
| 168 |
+
|
| 169 |
+
[More Information Needed]
|
| 170 |
+
|
| 171 |
+
## Citation [optional]
|
| 172 |
+
|
| 173 |
+
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
| 174 |
+
|
| 175 |
+
**BibTeX:**
|
| 176 |
+
|
| 177 |
+
[More Information Needed]
|
| 178 |
+
|
| 179 |
+
**APA:**
|
| 180 |
+
|
| 181 |
+
[More Information Needed]
|
| 182 |
+
|
| 183 |
+
## Glossary [optional]
|
| 184 |
+
|
| 185 |
+
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 186 |
+
|
| 187 |
+
[More Information Needed]
|
| 188 |
+
|
| 189 |
+
## More Information [optional]
|
| 190 |
+
|
| 191 |
+
[More Information Needed]
|
| 192 |
+
|
| 193 |
+
## Model Card Authors [optional]
|
| 194 |
+
|
| 195 |
+
[More Information Needed]
|
| 196 |
+
|
| 197 |
+
## Model Card Contact
|
| 198 |
+
|
| 199 |
+
[More Information Needed]
|
| 200 |
+
### Framework versions
|
| 201 |
+
|
| 202 |
+
- PEFT 0.14.0
|
checkpoint-1300/adapter_config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "microsoft/Phi-3-mini-4k-instruct",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"eva_config": null,
|
| 7 |
+
"exclude_modules": null,
|
| 8 |
+
"fan_in_fan_out": false,
|
| 9 |
+
"inference_mode": true,
|
| 10 |
+
"init_lora_weights": true,
|
| 11 |
+
"layer_replication": null,
|
| 12 |
+
"layers_pattern": null,
|
| 13 |
+
"layers_to_transform": null,
|
| 14 |
+
"loftq_config": {},
|
| 15 |
+
"lora_alpha": 16,
|
| 16 |
+
"lora_bias": false,
|
| 17 |
+
"lora_dropout": 0.0,
|
| 18 |
+
"megatron_config": null,
|
| 19 |
+
"megatron_core": "megatron.core",
|
| 20 |
+
"modules_to_save": null,
|
| 21 |
+
"peft_type": "LORA",
|
| 22 |
+
"r": 8,
|
| 23 |
+
"rank_pattern": {},
|
| 24 |
+
"revision": null,
|
| 25 |
+
"target_modules": [
|
| 26 |
+
"qkv_proj",
|
| 27 |
+
"gate_up_proj",
|
| 28 |
+
"down_proj",
|
| 29 |
+
"o_proj"
|
| 30 |
+
],
|
| 31 |
+
"task_type": "CAUSAL_LM",
|
| 32 |
+
"use_dora": false,
|
| 33 |
+
"use_rslora": false
|
| 34 |
+
}
|
checkpoint-1300/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7759af295987bc20f5132a0d2fd3c4c7c1a8c12c0c2a94f19b918babaadb7082
|
| 3 |
+
size 25200088
|
checkpoint-1300/global_step1299/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3893de4a8a1eb5e4a3a466b8abc65364db354fbda6141afa12f37d6c9cada7a
|
| 3 |
+
size 18881328
|
checkpoint-1300/global_step1299/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d04bdce43ee547363e131c34fcd14d13f914735ac6112beba832e4bf69d7c0ee
|
| 3 |
+
size 18881328
|
checkpoint-1300/global_step1299/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d57b33d1990cbfd9eee540d8466a3da9d9dba88699c4a0f2c6e49bc419dc272b
|
| 3 |
+
size 18881328
|
checkpoint-1300/global_step1299/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b97275f404e4c838e4dfa0e2777903f145ad61c53ef1684f114d948413dbd579
|
| 3 |
+
size 18881392
|
checkpoint-1300/global_step1299/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:71b846084a92cd59c95a9d198f4b129a0bda0356145bbb402567aaed500bb7a7
|
| 3 |
+
size 18881392
|
checkpoint-1300/global_step1299/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f92ad35b3dc8b311a02cd48e468d02f2ab08eeac48378d52204552706c7a24bc
|
| 3 |
+
size 18881392
|
checkpoint-1300/global_step1299/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:460133a896e4ea954bd39955d3679d217119add9c7586d5eeb19b3b7137ddc0f
|
| 3 |
+
size 18881392
|
checkpoint-1300/global_step1299/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9d520c9d0fbc57731406c4874b24bfe31440be1dd62eae50d9b1f8f316d89225
|
| 3 |
+
size 18881392
|
checkpoint-1300/global_step1299/mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b584be79fd902b3166d0eb6aa59f04ba80c66e9d8c36f61aa0f66b7a5846032
|
| 3 |
+
size 25379244
|
checkpoint-1300/latest
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
global_step1299
|
checkpoint-1300/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5eacda5f6af004ceddf64d89ee40cce618987575b41430389036f78b0db415ec
|
| 3 |
+
size 15984
|
checkpoint-1300/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:10cd1277d5978a7a496a54083286447baee888f75aabbb743adaac428980677a
|
| 3 |
+
size 15984
|
checkpoint-1300/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f9307e69f0bb11f9cd540f9f088ea591057a04216bdb8c4c24de8ba0080b466c
|
| 3 |
+
size 15984
|
checkpoint-1300/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6472c2fd22159b76dfb66618bec1669d7fb11e86eb472fe5271bc92708db57f3
|
| 3 |
+
size 15984
|
checkpoint-1300/rng_state_4.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8d5b4473cff161879a2b2e575d6bae51f67f85c4e702c7b941dc63776d0414ee
|
| 3 |
+
size 15984
|
checkpoint-1300/rng_state_5.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4e9b15ca6b8ae84725f6f0939a028bb92d50a69063e4b473b9f3100ae59e12c7
|
| 3 |
+
size 15984
|
checkpoint-1300/rng_state_6.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:097d7f11b98c5ed8f2b4ec8da21d513d95d6c3ff7b003e5000377430332f269a
|
| 3 |
+
size 15984
|
checkpoint-1300/rng_state_7.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e91bdab34024ebb46ea9609cfe31f3e29bef7962bbd9c877e3f53ed405be2375
|
| 3 |
+
size 15984
|
checkpoint-1300/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:915653f0579674c5a1f36a8b27b03dc55d5c94df5ebfb576cfff94db44ec6754
|
| 3 |
+
size 1064
|
checkpoint-1300/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|end|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "<|endoftext|>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<unk>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
checkpoint-1300/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1300/tokenizer_config.json
ADDED
|
@@ -0,0 +1,133 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "<unk>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"2": {
|
| 23 |
+
"content": "</s>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": false,
|
| 26 |
+
"rstrip": true,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": false
|
| 29 |
+
},
|
| 30 |
+
"32000": {
|
| 31 |
+
"content": "<|endoftext|>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false,
|
| 36 |
+
"special": true
|
| 37 |
+
},
|
| 38 |
+
"32001": {
|
| 39 |
+
"content": "<|assistant|>",
|
| 40 |
+
"lstrip": false,
|
| 41 |
+
"normalized": false,
|
| 42 |
+
"rstrip": true,
|
| 43 |
+
"single_word": false,
|
| 44 |
+
"special": true
|
| 45 |
+
},
|
| 46 |
+
"32002": {
|
| 47 |
+
"content": "<|placeholder1|>",
|
| 48 |
+
"lstrip": false,
|
| 49 |
+
"normalized": false,
|
| 50 |
+
"rstrip": true,
|
| 51 |
+
"single_word": false,
|
| 52 |
+
"special": true
|
| 53 |
+
},
|
| 54 |
+
"32003": {
|
| 55 |
+
"content": "<|placeholder2|>",
|
| 56 |
+
"lstrip": false,
|
| 57 |
+
"normalized": false,
|
| 58 |
+
"rstrip": true,
|
| 59 |
+
"single_word": false,
|
| 60 |
+
"special": true
|
| 61 |
+
},
|
| 62 |
+
"32004": {
|
| 63 |
+
"content": "<|placeholder3|>",
|
| 64 |
+
"lstrip": false,
|
| 65 |
+
"normalized": false,
|
| 66 |
+
"rstrip": true,
|
| 67 |
+
"single_word": false,
|
| 68 |
+
"special": true
|
| 69 |
+
},
|
| 70 |
+
"32005": {
|
| 71 |
+
"content": "<|placeholder4|>",
|
| 72 |
+
"lstrip": false,
|
| 73 |
+
"normalized": false,
|
| 74 |
+
"rstrip": true,
|
| 75 |
+
"single_word": false,
|
| 76 |
+
"special": true
|
| 77 |
+
},
|
| 78 |
+
"32006": {
|
| 79 |
+
"content": "<|system|>",
|
| 80 |
+
"lstrip": false,
|
| 81 |
+
"normalized": false,
|
| 82 |
+
"rstrip": true,
|
| 83 |
+
"single_word": false,
|
| 84 |
+
"special": true
|
| 85 |
+
},
|
| 86 |
+
"32007": {
|
| 87 |
+
"content": "<|end|>",
|
| 88 |
+
"lstrip": false,
|
| 89 |
+
"normalized": false,
|
| 90 |
+
"rstrip": false,
|
| 91 |
+
"single_word": false,
|
| 92 |
+
"special": true
|
| 93 |
+
},
|
| 94 |
+
"32008": {
|
| 95 |
+
"content": "<|placeholder5|>",
|
| 96 |
+
"lstrip": false,
|
| 97 |
+
"normalized": false,
|
| 98 |
+
"rstrip": true,
|
| 99 |
+
"single_word": false,
|
| 100 |
+
"special": true
|
| 101 |
+
},
|
| 102 |
+
"32009": {
|
| 103 |
+
"content": "<|placeholder6|>",
|
| 104 |
+
"lstrip": false,
|
| 105 |
+
"normalized": false,
|
| 106 |
+
"rstrip": true,
|
| 107 |
+
"single_word": false,
|
| 108 |
+
"special": true
|
| 109 |
+
},
|
| 110 |
+
"32010": {
|
| 111 |
+
"content": "<|user|>",
|
| 112 |
+
"lstrip": false,
|
| 113 |
+
"normalized": false,
|
| 114 |
+
"rstrip": true,
|
| 115 |
+
"single_word": false,
|
| 116 |
+
"special": true
|
| 117 |
+
}
|
| 118 |
+
},
|
| 119 |
+
"bos_token": "<s>",
|
| 120 |
+
"chat_template": "{% set system_message = 'You are a helpful AI assistant.' %}{% if messages[0]['role'] == 'system' %}{% set system_message = messages[0]['content'] %}{% endif %}{% if system_message is defined %}{{ '<s>' + '<|system|>\n' + system_message + '<|end|>\n' }}{% endif %}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '<|user|>\n' + content + '<|end|>\n<|assistant|>\n' }}{% elif message['role'] == 'assistant' %}{{ content + '<|end|>' + '\n' }}{% endif %}{% endfor %}",
|
| 121 |
+
"clean_up_tokenization_spaces": false,
|
| 122 |
+
"eos_token": "<|end|>",
|
| 123 |
+
"extra_special_tokens": {},
|
| 124 |
+
"legacy": false,
|
| 125 |
+
"model_max_length": 4096,
|
| 126 |
+
"pad_token": "<|endoftext|>",
|
| 127 |
+
"padding_side": "right",
|
| 128 |
+
"sp_model_kwargs": {},
|
| 129 |
+
"split_special_tokens": false,
|
| 130 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 131 |
+
"unk_token": "<unk>",
|
| 132 |
+
"use_default_system_prompt": false
|
| 133 |
+
}
|
checkpoint-1300/trainer_state.json
ADDED
|
@@ -0,0 +1,2399 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.1168886978942845,
|
| 5 |
+
"eval_steps": 50,
|
| 6 |
+
"global_step": 1300,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008594757198109154,
|
| 13 |
+
"grad_norm": 0.05934199318289757,
|
| 14 |
+
"learning_rate": 4.999451708687114e-06,
|
| 15 |
+
"logits/chosen": 14.762972831726074,
|
| 16 |
+
"logits/rejected": 15.199728012084961,
|
| 17 |
+
"logps/chosen": -0.3259914815425873,
|
| 18 |
+
"logps/rejected": -0.34297481179237366,
|
| 19 |
+
"loss": 0.9377,
|
| 20 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 21 |
+
"rewards/chosen": -0.4889872074127197,
|
| 22 |
+
"rewards/margins": 0.02547495998442173,
|
| 23 |
+
"rewards/rejected": -0.5144621729850769,
|
| 24 |
+
"step": 10
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.017189514396218308,
|
| 28 |
+
"grad_norm": 0.06342790275812149,
|
| 29 |
+
"learning_rate": 4.997807075247147e-06,
|
| 30 |
+
"logits/chosen": 14.351249694824219,
|
| 31 |
+
"logits/rejected": 15.068448066711426,
|
| 32 |
+
"logps/chosen": -0.2809392511844635,
|
| 33 |
+
"logps/rejected": -0.3711296617984772,
|
| 34 |
+
"loss": 0.9352,
|
| 35 |
+
"rewards/accuracies": 0.574999988079071,
|
| 36 |
+
"rewards/chosen": -0.42140883207321167,
|
| 37 |
+
"rewards/margins": 0.1352856159210205,
|
| 38 |
+
"rewards/rejected": -0.5566944479942322,
|
| 39 |
+
"step": 20
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.02578427159432746,
|
| 43 |
+
"grad_norm": 0.053961098194122314,
|
| 44 |
+
"learning_rate": 4.9950668210706795e-06,
|
| 45 |
+
"logits/chosen": 14.636960983276367,
|
| 46 |
+
"logits/rejected": 15.265243530273438,
|
| 47 |
+
"logps/chosen": -0.2820780873298645,
|
| 48 |
+
"logps/rejected": -0.34024301171302795,
|
| 49 |
+
"loss": 0.9351,
|
| 50 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 51 |
+
"rewards/chosen": -0.42311716079711914,
|
| 52 |
+
"rewards/margins": 0.08724743127822876,
|
| 53 |
+
"rewards/rejected": -0.5103646516799927,
|
| 54 |
+
"step": 30
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.034379028792436615,
|
| 58 |
+
"grad_norm": 0.13506193459033966,
|
| 59 |
+
"learning_rate": 4.9912321481237616e-06,
|
| 60 |
+
"logits/chosen": 14.4556884765625,
|
| 61 |
+
"logits/rejected": 15.048967361450195,
|
| 62 |
+
"logps/chosen": -0.2897028625011444,
|
| 63 |
+
"logps/rejected": -0.34129124879837036,
|
| 64 |
+
"loss": 0.922,
|
| 65 |
+
"rewards/accuracies": 0.44999998807907104,
|
| 66 |
+
"rewards/chosen": -0.43455424904823303,
|
| 67 |
+
"rewards/margins": 0.07738252729177475,
|
| 68 |
+
"rewards/rejected": -0.5119368433952332,
|
| 69 |
+
"step": 40
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.042973785990545764,
|
| 73 |
+
"grad_norm": 0.05230574309825897,
|
| 74 |
+
"learning_rate": 4.986304738420684e-06,
|
| 75 |
+
"logits/chosen": 14.628789901733398,
|
| 76 |
+
"logits/rejected": 15.307828903198242,
|
| 77 |
+
"logps/chosen": -0.28786614537239075,
|
| 78 |
+
"logps/rejected": -0.3513876795768738,
|
| 79 |
+
"loss": 0.9201,
|
| 80 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 81 |
+
"rewards/chosen": -0.4317992329597473,
|
| 82 |
+
"rewards/margins": 0.09528233855962753,
|
| 83 |
+
"rewards/rejected": -0.5270815491676331,
|
| 84 |
+
"step": 50
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.042973785990545764,
|
| 88 |
+
"eval_logits/chosen": 14.234943389892578,
|
| 89 |
+
"eval_logits/rejected": 15.258601188659668,
|
| 90 |
+
"eval_logps/chosen": -0.2844341993331909,
|
| 91 |
+
"eval_logps/rejected": -0.3695394694805145,
|
| 92 |
+
"eval_loss": 0.9226060509681702,
|
| 93 |
+
"eval_rewards/accuracies": 0.5157894492149353,
|
| 94 |
+
"eval_rewards/chosen": -0.42665132880210876,
|
| 95 |
+
"eval_rewards/margins": 0.1276579648256302,
|
| 96 |
+
"eval_rewards/rejected": -0.5543092489242554,
|
| 97 |
+
"eval_runtime": 25.9356,
|
| 98 |
+
"eval_samples_per_second": 29.033,
|
| 99 |
+
"eval_steps_per_second": 3.663,
|
| 100 |
+
"step": 50
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.05156854318865492,
|
| 104 |
+
"grad_norm": 0.09328428655862808,
|
| 105 |
+
"learning_rate": 4.980286753286196e-06,
|
| 106 |
+
"logits/chosen": 14.35963249206543,
|
| 107 |
+
"logits/rejected": 15.055354118347168,
|
| 108 |
+
"logps/chosen": -0.27534741163253784,
|
| 109 |
+
"logps/rejected": -0.33098170161247253,
|
| 110 |
+
"loss": 0.9356,
|
| 111 |
+
"rewards/accuracies": 0.512499988079071,
|
| 112 |
+
"rewards/chosen": -0.4130210876464844,
|
| 113 |
+
"rewards/margins": 0.08345144242048264,
|
| 114 |
+
"rewards/rejected": -0.4964725375175476,
|
| 115 |
+
"step": 60
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.060163300386764075,
|
| 119 |
+
"grad_norm": 0.06518550217151642,
|
| 120 |
+
"learning_rate": 4.973180832407471e-06,
|
| 121 |
+
"logits/chosen": 14.599525451660156,
|
| 122 |
+
"logits/rejected": 14.825297355651855,
|
| 123 |
+
"logps/chosen": -0.2708163857460022,
|
| 124 |
+
"logps/rejected": -0.3305850923061371,
|
| 125 |
+
"loss": 0.9257,
|
| 126 |
+
"rewards/accuracies": 0.550000011920929,
|
| 127 |
+
"rewards/chosen": -0.4062245786190033,
|
| 128 |
+
"rewards/margins": 0.08965305984020233,
|
| 129 |
+
"rewards/rejected": -0.4958776533603668,
|
| 130 |
+
"step": 70
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.06875805758487323,
|
| 134 |
+
"grad_norm": 0.07543154805898666,
|
| 135 |
+
"learning_rate": 4.964990092676263e-06,
|
| 136 |
+
"logits/chosen": 14.947430610656738,
|
| 137 |
+
"logits/rejected": 15.093690872192383,
|
| 138 |
+
"logps/chosen": -0.2602943778038025,
|
| 139 |
+
"logps/rejected": -0.31820863485336304,
|
| 140 |
+
"loss": 0.9168,
|
| 141 |
+
"rewards/accuracies": 0.5,
|
| 142 |
+
"rewards/chosen": -0.39044153690338135,
|
| 143 |
+
"rewards/margins": 0.08687138557434082,
|
| 144 |
+
"rewards/rejected": -0.47731298208236694,
|
| 145 |
+
"step": 80
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.07735281478298238,
|
| 149 |
+
"grad_norm": 0.06628195196390152,
|
| 150 |
+
"learning_rate": 4.9557181268217225e-06,
|
| 151 |
+
"logits/chosen": 14.43529987335205,
|
| 152 |
+
"logits/rejected": 14.750699043273926,
|
| 153 |
+
"logps/chosen": -0.2884291708469391,
|
| 154 |
+
"logps/rejected": -0.34193652868270874,
|
| 155 |
+
"loss": 0.9273,
|
| 156 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 157 |
+
"rewards/chosen": -0.43264374136924744,
|
| 158 |
+
"rewards/margins": 0.08026103675365448,
|
| 159 |
+
"rewards/rejected": -0.5129047632217407,
|
| 160 |
+
"step": 90
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.08594757198109153,
|
| 164 |
+
"grad_norm": 0.08684897422790527,
|
| 165 |
+
"learning_rate": 4.9453690018345144e-06,
|
| 166 |
+
"logits/chosen": 13.573002815246582,
|
| 167 |
+
"logits/rejected": 14.441877365112305,
|
| 168 |
+
"logps/chosen": -0.2569890320301056,
|
| 169 |
+
"logps/rejected": -0.37049269676208496,
|
| 170 |
+
"loss": 0.9009,
|
| 171 |
+
"rewards/accuracies": 0.574999988079071,
|
| 172 |
+
"rewards/chosen": -0.3854835629463196,
|
| 173 |
+
"rewards/margins": 0.17025551199913025,
|
| 174 |
+
"rewards/rejected": -0.5557390451431274,
|
| 175 |
+
"step": 100
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.08594757198109153,
|
| 179 |
+
"eval_logits/chosen": 14.026633262634277,
|
| 180 |
+
"eval_logits/rejected": 15.08835220336914,
|
| 181 |
+
"eval_logps/chosen": -0.2761566936969757,
|
| 182 |
+
"eval_logps/rejected": -0.3717801570892334,
|
| 183 |
+
"eval_loss": 0.9138591885566711,
|
| 184 |
+
"eval_rewards/accuracies": 0.5368421077728271,
|
| 185 |
+
"eval_rewards/chosen": -0.41423505544662476,
|
| 186 |
+
"eval_rewards/margins": 0.1434352546930313,
|
| 187 |
+
"eval_rewards/rejected": -0.5576702952384949,
|
| 188 |
+
"eval_runtime": 25.3996,
|
| 189 |
+
"eval_samples_per_second": 29.646,
|
| 190 |
+
"eval_steps_per_second": 3.74,
|
| 191 |
+
"step": 100
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.09454232917920069,
|
| 195 |
+
"grad_norm": 0.08046824485063553,
|
| 196 |
+
"learning_rate": 4.933947257182901e-06,
|
| 197 |
+
"logits/chosen": 14.500630378723145,
|
| 198 |
+
"logits/rejected": 14.831761360168457,
|
| 199 |
+
"logps/chosen": -0.30049553513526917,
|
| 200 |
+
"logps/rejected": -0.3315966725349426,
|
| 201 |
+
"loss": 0.916,
|
| 202 |
+
"rewards/accuracies": 0.4625000059604645,
|
| 203 |
+
"rewards/chosen": -0.45074325799942017,
|
| 204 |
+
"rewards/margins": 0.04665176197886467,
|
| 205 |
+
"rewards/rejected": -0.49739497900009155,
|
| 206 |
+
"step": 110
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.10313708637730984,
|
| 210 |
+
"grad_norm": 0.12244562804698944,
|
| 211 |
+
"learning_rate": 4.921457902821578e-06,
|
| 212 |
+
"logits/chosen": 14.26713752746582,
|
| 213 |
+
"logits/rejected": 14.495455741882324,
|
| 214 |
+
"logps/chosen": -0.2670941650867462,
|
| 215 |
+
"logps/rejected": -0.32481229305267334,
|
| 216 |
+
"loss": 0.9167,
|
| 217 |
+
"rewards/accuracies": 0.550000011920929,
|
| 218 |
+
"rewards/chosen": -0.4006412625312805,
|
| 219 |
+
"rewards/margins": 0.08657723665237427,
|
| 220 |
+
"rewards/rejected": -0.4872184693813324,
|
| 221 |
+
"step": 120
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.11173184357541899,
|
| 225 |
+
"grad_norm": 0.1828213334083557,
|
| 226 |
+
"learning_rate": 4.907906416994146e-06,
|
| 227 |
+
"logits/chosen": 14.009546279907227,
|
| 228 |
+
"logits/rejected": 14.297094345092773,
|
| 229 |
+
"logps/chosen": -0.27995598316192627,
|
| 230 |
+
"logps/rejected": -0.3530685007572174,
|
| 231 |
+
"loss": 0.9087,
|
| 232 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 233 |
+
"rewards/chosen": -0.419933944940567,
|
| 234 |
+
"rewards/margins": 0.10966875404119492,
|
| 235 |
+
"rewards/rejected": -0.5296027660369873,
|
| 236 |
+
"step": 130
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.12032660077352815,
|
| 240 |
+
"grad_norm": 0.10407563298940659,
|
| 241 |
+
"learning_rate": 4.893298743830168e-06,
|
| 242 |
+
"logits/chosen": 13.689155578613281,
|
| 243 |
+
"logits/rejected": 14.1933012008667,
|
| 244 |
+
"logps/chosen": -0.25955715775489807,
|
| 245 |
+
"logps/rejected": -0.3815004229545593,
|
| 246 |
+
"loss": 0.9053,
|
| 247 |
+
"rewards/accuracies": 0.612500011920929,
|
| 248 |
+
"rewards/chosen": -0.3893357217311859,
|
| 249 |
+
"rewards/margins": 0.18291489779949188,
|
| 250 |
+
"rewards/rejected": -0.5722506046295166,
|
| 251 |
+
"step": 140
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.1289213579716373,
|
| 255 |
+
"grad_norm": 0.10028588026762009,
|
| 256 |
+
"learning_rate": 4.8776412907378845e-06,
|
| 257 |
+
"logits/chosen": 12.851397514343262,
|
| 258 |
+
"logits/rejected": 13.509778022766113,
|
| 259 |
+
"logps/chosen": -0.23652991652488708,
|
| 260 |
+
"logps/rejected": -0.3720462918281555,
|
| 261 |
+
"loss": 0.8999,
|
| 262 |
+
"rewards/accuracies": 0.625,
|
| 263 |
+
"rewards/chosen": -0.3547949194908142,
|
| 264 |
+
"rewards/margins": 0.2032744586467743,
|
| 265 |
+
"rewards/rejected": -0.5580693483352661,
|
| 266 |
+
"step": 150
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.1289213579716373,
|
| 270 |
+
"eval_logits/chosen": 12.384929656982422,
|
| 271 |
+
"eval_logits/rejected": 13.672826766967773,
|
| 272 |
+
"eval_logps/chosen": -0.27857670187950134,
|
| 273 |
+
"eval_logps/rejected": -0.4014737904071808,
|
| 274 |
+
"eval_loss": 0.8956203460693359,
|
| 275 |
+
"eval_rewards/accuracies": 0.5684210658073425,
|
| 276 |
+
"eval_rewards/chosen": -0.4178650677204132,
|
| 277 |
+
"eval_rewards/margins": 0.18434564769268036,
|
| 278 |
+
"eval_rewards/rejected": -0.6022107601165771,
|
| 279 |
+
"eval_runtime": 25.4176,
|
| 280 |
+
"eval_samples_per_second": 29.625,
|
| 281 |
+
"eval_steps_per_second": 3.738,
|
| 282 |
+
"step": 150
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.13751611516974646,
|
| 286 |
+
"grad_norm": 0.12453093379735947,
|
| 287 |
+
"learning_rate": 4.860940925593703e-06,
|
| 288 |
+
"logits/chosen": 12.110003471374512,
|
| 289 |
+
"logits/rejected": 13.076980590820312,
|
| 290 |
+
"logps/chosen": -0.27192068099975586,
|
| 291 |
+
"logps/rejected": -0.3863692879676819,
|
| 292 |
+
"loss": 0.8907,
|
| 293 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 294 |
+
"rewards/chosen": -0.4078810214996338,
|
| 295 |
+
"rewards/margins": 0.1716729700565338,
|
| 296 |
+
"rewards/rejected": -0.5795539617538452,
|
| 297 |
+
"step": 160
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.1461108723678556,
|
| 301 |
+
"grad_norm": 0.17137788236141205,
|
| 302 |
+
"learning_rate": 4.84320497372973e-06,
|
| 303 |
+
"logits/chosen": 11.92918586730957,
|
| 304 |
+
"logits/rejected": 12.573629379272461,
|
| 305 |
+
"logps/chosen": -0.27472984790802,
|
| 306 |
+
"logps/rejected": -0.41249385476112366,
|
| 307 |
+
"loss": 0.8831,
|
| 308 |
+
"rewards/accuracies": 0.612500011920929,
|
| 309 |
+
"rewards/chosen": -0.41209474205970764,
|
| 310 |
+
"rewards/margins": 0.20664596557617188,
|
| 311 |
+
"rewards/rejected": -0.6187406778335571,
|
| 312 |
+
"step": 170
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.15470562956596476,
|
| 316 |
+
"grad_norm": 0.3904883861541748,
|
| 317 |
+
"learning_rate": 4.824441214720629e-06,
|
| 318 |
+
"logits/chosen": 11.182531356811523,
|
| 319 |
+
"logits/rejected": 12.176573753356934,
|
| 320 |
+
"logps/chosen": -0.2953718304634094,
|
| 321 |
+
"logps/rejected": -0.4208717942237854,
|
| 322 |
+
"loss": 0.8736,
|
| 323 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 324 |
+
"rewards/chosen": -0.4430577754974365,
|
| 325 |
+
"rewards/margins": 0.18824996054172516,
|
| 326 |
+
"rewards/rejected": -0.6313077211380005,
|
| 327 |
+
"step": 180
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.1633003867640739,
|
| 331 |
+
"grad_norm": 0.17574089765548706,
|
| 332 |
+
"learning_rate": 4.804657878971252e-06,
|
| 333 |
+
"logits/chosen": 10.119890213012695,
|
| 334 |
+
"logits/rejected": 11.05900764465332,
|
| 335 |
+
"logps/chosen": -0.29340866208076477,
|
| 336 |
+
"logps/rejected": -0.4555762708187103,
|
| 337 |
+
"loss": 0.884,
|
| 338 |
+
"rewards/accuracies": 0.625,
|
| 339 |
+
"rewards/chosen": -0.44011297821998596,
|
| 340 |
+
"rewards/margins": 0.24325144290924072,
|
| 341 |
+
"rewards/rejected": -0.6833644509315491,
|
| 342 |
+
"step": 190
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.17189514396218306,
|
| 346 |
+
"grad_norm": 0.2242884337902069,
|
| 347 |
+
"learning_rate": 4.783863644106502e-06,
|
| 348 |
+
"logits/chosen": 9.674784660339355,
|
| 349 |
+
"logits/rejected": 10.418611526489258,
|
| 350 |
+
"logps/chosen": -0.3504490852355957,
|
| 351 |
+
"logps/rejected": -0.5431731939315796,
|
| 352 |
+
"loss": 0.8419,
|
| 353 |
+
"rewards/accuracies": 0.612500011920929,
|
| 354 |
+
"rewards/chosen": -0.5256736278533936,
|
| 355 |
+
"rewards/margins": 0.2890861928462982,
|
| 356 |
+
"rewards/rejected": -0.8147598505020142,
|
| 357 |
+
"step": 200
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.17189514396218306,
|
| 361 |
+
"eval_logits/chosen": 7.944870471954346,
|
| 362 |
+
"eval_logits/rejected": 8.979729652404785,
|
| 363 |
+
"eval_logps/chosen": -0.33341673016548157,
|
| 364 |
+
"eval_logps/rejected": -0.5431775450706482,
|
| 365 |
+
"eval_loss": 0.8462886810302734,
|
| 366 |
+
"eval_rewards/accuracies": 0.6000000238418579,
|
| 367 |
+
"eval_rewards/chosen": -0.5001251101493835,
|
| 368 |
+
"eval_rewards/margins": 0.3146411180496216,
|
| 369 |
+
"eval_rewards/rejected": -0.8147663474082947,
|
| 370 |
+
"eval_runtime": 25.419,
|
| 371 |
+
"eval_samples_per_second": 29.623,
|
| 372 |
+
"eval_steps_per_second": 3.737,
|
| 373 |
+
"step": 200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.18048990116029223,
|
| 377 |
+
"grad_norm": 0.32119837403297424,
|
| 378 |
+
"learning_rate": 4.762067631165049e-06,
|
| 379 |
+
"logits/chosen": 7.16138219833374,
|
| 380 |
+
"logits/rejected": 8.43680477142334,
|
| 381 |
+
"logps/chosen": -0.36649250984191895,
|
| 382 |
+
"logps/rejected": -0.5420924425125122,
|
| 383 |
+
"loss": 0.8187,
|
| 384 |
+
"rewards/accuracies": 0.612500011920929,
|
| 385 |
+
"rewards/chosen": -0.5497387647628784,
|
| 386 |
+
"rewards/margins": 0.2633998692035675,
|
| 387 |
+
"rewards/rejected": -0.8131386041641235,
|
| 388 |
+
"step": 210
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.18908465835840138,
|
| 392 |
+
"grad_norm": 0.48516562581062317,
|
| 393 |
+
"learning_rate": 4.7392794005985324e-06,
|
| 394 |
+
"logits/chosen": 4.770083427429199,
|
| 395 |
+
"logits/rejected": 5.710458278656006,
|
| 396 |
+
"logps/chosen": -0.34041497111320496,
|
| 397 |
+
"logps/rejected": -0.6309320330619812,
|
| 398 |
+
"loss": 0.8448,
|
| 399 |
+
"rewards/accuracies": 0.675000011920929,
|
| 400 |
+
"rewards/chosen": -0.510622501373291,
|
| 401 |
+
"rewards/margins": 0.4357755780220032,
|
| 402 |
+
"rewards/rejected": -0.9463980793952942,
|
| 403 |
+
"step": 220
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 0.19767941555651053,
|
| 407 |
+
"grad_norm": 0.29154208302497864,
|
| 408 |
+
"learning_rate": 4.715508948078037e-06,
|
| 409 |
+
"logits/chosen": 5.168765068054199,
|
| 410 |
+
"logits/rejected": 5.421420574188232,
|
| 411 |
+
"logps/chosen": -0.3792352080345154,
|
| 412 |
+
"logps/rejected": -0.65748131275177,
|
| 413 |
+
"loss": 0.8066,
|
| 414 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 415 |
+
"rewards/chosen": -0.5688528418540955,
|
| 416 |
+
"rewards/margins": 0.41736921668052673,
|
| 417 |
+
"rewards/rejected": -0.986221969127655,
|
| 418 |
+
"step": 230
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.20627417275461968,
|
| 422 |
+
"grad_norm": 0.42973750829696655,
|
| 423 |
+
"learning_rate": 4.690766700109659e-06,
|
| 424 |
+
"logits/chosen": 4.204717636108398,
|
| 425 |
+
"logits/rejected": 3.706291913986206,
|
| 426 |
+
"logps/chosen": -0.39414530992507935,
|
| 427 |
+
"logps/rejected": -0.7194588780403137,
|
| 428 |
+
"loss": 0.7787,
|
| 429 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 430 |
+
"rewards/chosen": -0.5912179350852966,
|
| 431 |
+
"rewards/margins": 0.4879704415798187,
|
| 432 |
+
"rewards/rejected": -1.079188346862793,
|
| 433 |
+
"step": 240
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.21486892995272883,
|
| 437 |
+
"grad_norm": 0.5244571566581726,
|
| 438 |
+
"learning_rate": 4.665063509461098e-06,
|
| 439 |
+
"logits/chosen": 3.335484743118286,
|
| 440 |
+
"logits/rejected": 3.3176345825195312,
|
| 441 |
+
"logps/chosen": -0.4493131637573242,
|
| 442 |
+
"logps/rejected": -0.8293434381484985,
|
| 443 |
+
"loss": 0.7776,
|
| 444 |
+
"rewards/accuracies": 0.625,
|
| 445 |
+
"rewards/chosen": -0.6739697456359863,
|
| 446 |
+
"rewards/margins": 0.5700454115867615,
|
| 447 |
+
"rewards/rejected": -1.244015097618103,
|
| 448 |
+
"step": 250
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.21486892995272883,
|
| 452 |
+
"eval_logits/chosen": 2.590949058532715,
|
| 453 |
+
"eval_logits/rejected": 2.2929749488830566,
|
| 454 |
+
"eval_logps/chosen": -0.48714593052864075,
|
| 455 |
+
"eval_logps/rejected": -0.9267774224281311,
|
| 456 |
+
"eval_loss": 0.7469337582588196,
|
| 457 |
+
"eval_rewards/accuracies": 0.6526315808296204,
|
| 458 |
+
"eval_rewards/chosen": -0.7307189106941223,
|
| 459 |
+
"eval_rewards/margins": 0.659447193145752,
|
| 460 |
+
"eval_rewards/rejected": -1.390166163444519,
|
| 461 |
+
"eval_runtime": 25.3944,
|
| 462 |
+
"eval_samples_per_second": 29.652,
|
| 463 |
+
"eval_steps_per_second": 3.741,
|
| 464 |
+
"step": 250
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.22346368715083798,
|
| 468 |
+
"grad_norm": 0.39347293972969055,
|
| 469 |
+
"learning_rate": 4.638410650401267e-06,
|
| 470 |
+
"logits/chosen": 2.2975668907165527,
|
| 471 |
+
"logits/rejected": 1.2855035066604614,
|
| 472 |
+
"logps/chosen": -0.5228341817855835,
|
| 473 |
+
"logps/rejected": -1.00227952003479,
|
| 474 |
+
"loss": 0.6981,
|
| 475 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 476 |
+
"rewards/chosen": -0.78425133228302,
|
| 477 |
+
"rewards/margins": 0.7191681265830994,
|
| 478 |
+
"rewards/rejected": -1.5034195184707642,
|
| 479 |
+
"step": 260
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.23205844434894715,
|
| 483 |
+
"grad_norm": 0.69575434923172,
|
| 484 |
+
"learning_rate": 4.610819813755038e-06,
|
| 485 |
+
"logits/chosen": 2.8782780170440674,
|
| 486 |
+
"logits/rejected": 1.9394336938858032,
|
| 487 |
+
"logps/chosen": -0.4982885718345642,
|
| 488 |
+
"logps/rejected": -1.035541296005249,
|
| 489 |
+
"loss": 0.7174,
|
| 490 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 491 |
+
"rewards/chosen": -0.7474328875541687,
|
| 492 |
+
"rewards/margins": 0.8058789372444153,
|
| 493 |
+
"rewards/rejected": -1.5533119440078735,
|
| 494 |
+
"step": 270
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.2406532015470563,
|
| 498 |
+
"grad_norm": 0.7858326435089111,
|
| 499 |
+
"learning_rate": 4.582303101775249e-06,
|
| 500 |
+
"logits/chosen": 2.710908889770508,
|
| 501 |
+
"logits/rejected": 1.6444288492202759,
|
| 502 |
+
"logps/chosen": -0.600068211555481,
|
| 503 |
+
"logps/rejected": -1.1271780729293823,
|
| 504 |
+
"loss": 0.6972,
|
| 505 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 506 |
+
"rewards/chosen": -0.9001023173332214,
|
| 507 |
+
"rewards/margins": 0.7906648516654968,
|
| 508 |
+
"rewards/rejected": -1.6907672882080078,
|
| 509 |
+
"step": 280
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.24924795874516545,
|
| 513 |
+
"grad_norm": 0.7384620904922485,
|
| 514 |
+
"learning_rate": 4.55287302283426e-06,
|
| 515 |
+
"logits/chosen": 1.5841500759124756,
|
| 516 |
+
"logits/rejected": 0.640514612197876,
|
| 517 |
+
"logps/chosen": -0.6465060710906982,
|
| 518 |
+
"logps/rejected": -1.4245095252990723,
|
| 519 |
+
"loss": 0.6192,
|
| 520 |
+
"rewards/accuracies": 0.612500011920929,
|
| 521 |
+
"rewards/chosen": -0.9697591066360474,
|
| 522 |
+
"rewards/margins": 1.1670053005218506,
|
| 523 |
+
"rewards/rejected": -2.1367642879486084,
|
| 524 |
+
"step": 290
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.2578427159432746,
|
| 528 |
+
"grad_norm": 0.8262321352958679,
|
| 529 |
+
"learning_rate": 4.522542485937369e-06,
|
| 530 |
+
"logits/chosen": 1.7300422191619873,
|
| 531 |
+
"logits/rejected": 0.7782856225967407,
|
| 532 |
+
"logps/chosen": -0.7083590626716614,
|
| 533 |
+
"logps/rejected": -1.6742557287216187,
|
| 534 |
+
"loss": 0.5721,
|
| 535 |
+
"rewards/accuracies": 0.612500011920929,
|
| 536 |
+
"rewards/chosen": -1.062538504600525,
|
| 537 |
+
"rewards/margins": 1.4488452672958374,
|
| 538 |
+
"rewards/rejected": -2.511383533477783,
|
| 539 |
+
"step": 300
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 0.2578427159432746,
|
| 543 |
+
"eval_logits/chosen": 1.3559931516647339,
|
| 544 |
+
"eval_logits/rejected": 0.6592276096343994,
|
| 545 |
+
"eval_logps/chosen": -0.7815767526626587,
|
| 546 |
+
"eval_logps/rejected": -2.1176154613494873,
|
| 547 |
+
"eval_loss": 0.5730626583099365,
|
| 548 |
+
"eval_rewards/accuracies": 0.7052631378173828,
|
| 549 |
+
"eval_rewards/chosen": -1.1723653078079224,
|
| 550 |
+
"eval_rewards/margins": 2.0040581226348877,
|
| 551 |
+
"eval_rewards/rejected": -3.1764233112335205,
|
| 552 |
+
"eval_runtime": 25.539,
|
| 553 |
+
"eval_samples_per_second": 29.484,
|
| 554 |
+
"eval_steps_per_second": 3.72,
|
| 555 |
+
"step": 300
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.2664374731413838,
|
| 559 |
+
"grad_norm": 0.8472572565078735,
|
| 560 |
+
"learning_rate": 4.491324795060491e-06,
|
| 561 |
+
"logits/chosen": 1.4461088180541992,
|
| 562 |
+
"logits/rejected": 0.49669915437698364,
|
| 563 |
+
"logps/chosen": -0.7694377899169922,
|
| 564 |
+
"logps/rejected": -2.362783432006836,
|
| 565 |
+
"loss": 0.5091,
|
| 566 |
+
"rewards/accuracies": 0.762499988079071,
|
| 567 |
+
"rewards/chosen": -1.1541565656661987,
|
| 568 |
+
"rewards/margins": 2.390018939971924,
|
| 569 |
+
"rewards/rejected": -3.544174909591675,
|
| 570 |
+
"step": 310
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.2750322303394929,
|
| 574 |
+
"grad_norm": 0.41847530007362366,
|
| 575 |
+
"learning_rate": 4.4592336433146e-06,
|
| 576 |
+
"logits/chosen": 2.172646999359131,
|
| 577 |
+
"logits/rejected": 1.0526962280273438,
|
| 578 |
+
"logps/chosen": -0.7410945296287537,
|
| 579 |
+
"logps/rejected": -1.9158353805541992,
|
| 580 |
+
"loss": 0.5352,
|
| 581 |
+
"rewards/accuracies": 0.675000011920929,
|
| 582 |
+
"rewards/chosen": -1.1116416454315186,
|
| 583 |
+
"rewards/margins": 1.7621114253997803,
|
| 584 |
+
"rewards/rejected": -2.873753070831299,
|
| 585 |
+
"step": 320
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.28362698753760207,
|
| 589 |
+
"grad_norm": 1.7422096729278564,
|
| 590 |
+
"learning_rate": 4.426283106939474e-06,
|
| 591 |
+
"logits/chosen": 2.611234188079834,
|
| 592 |
+
"logits/rejected": 1.7068111896514893,
|
| 593 |
+
"logps/chosen": -0.8319486379623413,
|
| 594 |
+
"logps/rejected": -2.32024884223938,
|
| 595 |
+
"loss": 0.5397,
|
| 596 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 597 |
+
"rewards/chosen": -1.2479230165481567,
|
| 598 |
+
"rewards/margins": 2.232450008392334,
|
| 599 |
+
"rewards/rejected": -3.480372905731201,
|
| 600 |
+
"step": 330
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.2922217447357112,
|
| 604 |
+
"grad_norm": 0.8699240684509277,
|
| 605 |
+
"learning_rate": 4.3924876391293915e-06,
|
| 606 |
+
"logits/chosen": 1.996747612953186,
|
| 607 |
+
"logits/rejected": 1.1473515033721924,
|
| 608 |
+
"logps/chosen": -0.8445833921432495,
|
| 609 |
+
"logps/rejected": -2.675687551498413,
|
| 610 |
+
"loss": 0.4817,
|
| 611 |
+
"rewards/accuracies": 0.699999988079071,
|
| 612 |
+
"rewards/chosen": -1.2668750286102295,
|
| 613 |
+
"rewards/margins": 2.7466559410095215,
|
| 614 |
+
"rewards/rejected": -4.01353120803833,
|
| 615 |
+
"step": 340
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.30081650193382037,
|
| 619 |
+
"grad_norm": 2.089289426803589,
|
| 620 |
+
"learning_rate": 4.357862063693486e-06,
|
| 621 |
+
"logits/chosen": 1.7134803533554077,
|
| 622 |
+
"logits/rejected": 1.3000510931015015,
|
| 623 |
+
"logps/chosen": -0.8976927995681763,
|
| 624 |
+
"logps/rejected": -2.1593873500823975,
|
| 625 |
+
"loss": 0.5098,
|
| 626 |
+
"rewards/accuracies": 0.574999988079071,
|
| 627 |
+
"rewards/chosen": -1.3465392589569092,
|
| 628 |
+
"rewards/margins": 1.8925418853759766,
|
| 629 |
+
"rewards/rejected": -3.2390809059143066,
|
| 630 |
+
"step": 350
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 0.30081650193382037,
|
| 634 |
+
"eval_logits/chosen": 1.6772903203964233,
|
| 635 |
+
"eval_logits/rejected": 1.2370609045028687,
|
| 636 |
+
"eval_logps/chosen": -0.9737761616706848,
|
| 637 |
+
"eval_logps/rejected": -3.1528680324554443,
|
| 638 |
+
"eval_loss": 0.5162621736526489,
|
| 639 |
+
"eval_rewards/accuracies": 0.7263157963752747,
|
| 640 |
+
"eval_rewards/chosen": -1.46066415309906,
|
| 641 |
+
"eval_rewards/margins": 3.2686376571655273,
|
| 642 |
+
"eval_rewards/rejected": -4.729301929473877,
|
| 643 |
+
"eval_runtime": 25.4163,
|
| 644 |
+
"eval_samples_per_second": 29.627,
|
| 645 |
+
"eval_steps_per_second": 3.738,
|
| 646 |
+
"step": 350
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.3094112591319295,
|
| 650 |
+
"grad_norm": 0.47079572081565857,
|
| 651 |
+
"learning_rate": 4.322421568553529e-06,
|
| 652 |
+
"logits/chosen": 1.9561872482299805,
|
| 653 |
+
"logits/rejected": 0.8960329294204712,
|
| 654 |
+
"logps/chosen": -0.9378088712692261,
|
| 655 |
+
"logps/rejected": -2.8065876960754395,
|
| 656 |
+
"loss": 0.5046,
|
| 657 |
+
"rewards/accuracies": 0.675000011920929,
|
| 658 |
+
"rewards/chosen": -1.4067132472991943,
|
| 659 |
+
"rewards/margins": 2.8031680583953857,
|
| 660 |
+
"rewards/rejected": -4.209881782531738,
|
| 661 |
+
"step": 360
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.31800601633003867,
|
| 665 |
+
"grad_norm": 0.6202365159988403,
|
| 666 |
+
"learning_rate": 4.286181699082008e-06,
|
| 667 |
+
"logits/chosen": 2.152726411819458,
|
| 668 |
+
"logits/rejected": 1.4309433698654175,
|
| 669 |
+
"logps/chosen": -1.007157564163208,
|
| 670 |
+
"logps/rejected": -3.3813462257385254,
|
| 671 |
+
"loss": 0.4526,
|
| 672 |
+
"rewards/accuracies": 0.800000011920929,
|
| 673 |
+
"rewards/chosen": -1.5107364654541016,
|
| 674 |
+
"rewards/margins": 3.561283588409424,
|
| 675 |
+
"rewards/rejected": -5.072019577026367,
|
| 676 |
+
"step": 370
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.3266007735281478,
|
| 680 |
+
"grad_norm": 1.080393671989441,
|
| 681 |
+
"learning_rate": 4.249158351283414e-06,
|
| 682 |
+
"logits/chosen": 1.7528371810913086,
|
| 683 |
+
"logits/rejected": 1.3293968439102173,
|
| 684 |
+
"logps/chosen": -1.0258004665374756,
|
| 685 |
+
"logps/rejected": -2.984057903289795,
|
| 686 |
+
"loss": 0.4879,
|
| 687 |
+
"rewards/accuracies": 0.675000011920929,
|
| 688 |
+
"rewards/chosen": -1.5387006998062134,
|
| 689 |
+
"rewards/margins": 2.9373860359191895,
|
| 690 |
+
"rewards/rejected": -4.476086616516113,
|
| 691 |
+
"step": 380
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.33519553072625696,
|
| 695 |
+
"grad_norm": 1.4520032405853271,
|
| 696 |
+
"learning_rate": 4.211367764821722e-06,
|
| 697 |
+
"logits/chosen": 3.061373233795166,
|
| 698 |
+
"logits/rejected": 2.0103466510772705,
|
| 699 |
+
"logps/chosen": -1.0191391706466675,
|
| 700 |
+
"logps/rejected": -2.9054081439971924,
|
| 701 |
+
"loss": 0.4776,
|
| 702 |
+
"rewards/accuracies": 0.625,
|
| 703 |
+
"rewards/chosen": -1.5287089347839355,
|
| 704 |
+
"rewards/margins": 2.8294031620025635,
|
| 705 |
+
"rewards/rejected": -4.358112335205078,
|
| 706 |
+
"step": 390
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.3437902879243661,
|
| 710 |
+
"grad_norm": 0.5479139089584351,
|
| 711 |
+
"learning_rate": 4.172826515897146e-06,
|
| 712 |
+
"logits/chosen": 2.8395092487335205,
|
| 713 |
+
"logits/rejected": 2.0935282707214355,
|
| 714 |
+
"logps/chosen": -1.0769506692886353,
|
| 715 |
+
"logps/rejected": -3.11635160446167,
|
| 716 |
+
"loss": 0.4686,
|
| 717 |
+
"rewards/accuracies": 0.6875,
|
| 718 |
+
"rewards/chosen": -1.6154258251190186,
|
| 719 |
+
"rewards/margins": 3.0591015815734863,
|
| 720 |
+
"rewards/rejected": -4.674527168273926,
|
| 721 |
+
"step": 400
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 0.3437902879243661,
|
| 725 |
+
"eval_logits/chosen": 2.5064592361450195,
|
| 726 |
+
"eval_logits/rejected": 2.108433485031128,
|
| 727 |
+
"eval_logps/chosen": -1.1957285404205322,
|
| 728 |
+
"eval_logps/rejected": -3.7678382396698,
|
| 729 |
+
"eval_loss": 0.46578800678253174,
|
| 730 |
+
"eval_rewards/accuracies": 0.7368420958518982,
|
| 731 |
+
"eval_rewards/chosen": -1.793592929840088,
|
| 732 |
+
"eval_rewards/margins": 3.8581647872924805,
|
| 733 |
+
"eval_rewards/rejected": -5.651757717132568,
|
| 734 |
+
"eval_runtime": 25.415,
|
| 735 |
+
"eval_samples_per_second": 29.628,
|
| 736 |
+
"eval_steps_per_second": 3.738,
|
| 737 |
+
"step": 400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.3523850451224753,
|
| 741 |
+
"grad_norm": 0.9966821670532227,
|
| 742 |
+
"learning_rate": 4.133551509975264e-06,
|
| 743 |
+
"logits/chosen": 2.6411917209625244,
|
| 744 |
+
"logits/rejected": 1.8634885549545288,
|
| 745 |
+
"logps/chosen": -1.0934125185012817,
|
| 746 |
+
"logps/rejected": -3.2207794189453125,
|
| 747 |
+
"loss": 0.4335,
|
| 748 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 749 |
+
"rewards/chosen": -1.6401188373565674,
|
| 750 |
+
"rewards/margins": 3.1910502910614014,
|
| 751 |
+
"rewards/rejected": -4.831169128417969,
|
| 752 |
+
"step": 410
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.36097980232058446,
|
| 756 |
+
"grad_norm": 0.6384722590446472,
|
| 757 |
+
"learning_rate": 4.093559974371725e-06,
|
| 758 |
+
"logits/chosen": 3.1368844509124756,
|
| 759 |
+
"logits/rejected": 2.3800251483917236,
|
| 760 |
+
"logps/chosen": -1.2108217477798462,
|
| 761 |
+
"logps/rejected": -3.484806537628174,
|
| 762 |
+
"loss": 0.4543,
|
| 763 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 764 |
+
"rewards/chosen": -1.816232681274414,
|
| 765 |
+
"rewards/margins": 3.4109771251678467,
|
| 766 |
+
"rewards/rejected": -5.227209568023682,
|
| 767 |
+
"step": 420
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.3695745595186936,
|
| 771 |
+
"grad_norm": 0.856741726398468,
|
| 772 |
+
"learning_rate": 4.052869450695776e-06,
|
| 773 |
+
"logits/chosen": 3.155728816986084,
|
| 774 |
+
"logits/rejected": 2.257838726043701,
|
| 775 |
+
"logps/chosen": -1.4214586019515991,
|
| 776 |
+
"logps/rejected": -4.186622619628906,
|
| 777 |
+
"loss": 0.4091,
|
| 778 |
+
"rewards/accuracies": 0.7749999761581421,
|
| 779 |
+
"rewards/chosen": -2.132187604904175,
|
| 780 |
+
"rewards/margins": 4.1477460861206055,
|
| 781 |
+
"rewards/rejected": -6.279933929443359,
|
| 782 |
+
"step": 430
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.37816931671680276,
|
| 786 |
+
"grad_norm": 1.3310774564743042,
|
| 787 |
+
"learning_rate": 4.011497787155938e-06,
|
| 788 |
+
"logits/chosen": 1.9942185878753662,
|
| 789 |
+
"logits/rejected": 1.6246827840805054,
|
| 790 |
+
"logps/chosen": -1.8575637340545654,
|
| 791 |
+
"logps/rejected": -4.5355329513549805,
|
| 792 |
+
"loss": 0.3995,
|
| 793 |
+
"rewards/accuracies": 0.862500011920929,
|
| 794 |
+
"rewards/chosen": -2.7863457202911377,
|
| 795 |
+
"rewards/margins": 4.016953945159912,
|
| 796 |
+
"rewards/rejected": -6.8032989501953125,
|
| 797 |
+
"step": 440
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 0.3867640739149119,
|
| 801 |
+
"grad_norm": 2.0849101543426514,
|
| 802 |
+
"learning_rate": 3.969463130731183e-06,
|
| 803 |
+
"logits/chosen": 2.406555652618408,
|
| 804 |
+
"logits/rejected": 2.0490009784698486,
|
| 805 |
+
"logps/chosen": -2.392570972442627,
|
| 806 |
+
"logps/rejected": -5.055584907531738,
|
| 807 |
+
"loss": 0.3671,
|
| 808 |
+
"rewards/accuracies": 0.887499988079071,
|
| 809 |
+
"rewards/chosen": -3.588855743408203,
|
| 810 |
+
"rewards/margins": 3.994520902633667,
|
| 811 |
+
"rewards/rejected": -7.583376884460449,
|
| 812 |
+
"step": 450
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.3867640739149119,
|
| 816 |
+
"eval_logits/chosen": 2.2324020862579346,
|
| 817 |
+
"eval_logits/rejected": 2.365755319595337,
|
| 818 |
+
"eval_logps/chosen": -2.736898422241211,
|
| 819 |
+
"eval_logps/rejected": -5.73967170715332,
|
| 820 |
+
"eval_loss": 0.3965117633342743,
|
| 821 |
+
"eval_rewards/accuracies": 0.8736842274665833,
|
| 822 |
+
"eval_rewards/chosen": -4.105347633361816,
|
| 823 |
+
"eval_rewards/margins": 4.504159927368164,
|
| 824 |
+
"eval_rewards/rejected": -8.60950756072998,
|
| 825 |
+
"eval_runtime": 25.428,
|
| 826 |
+
"eval_samples_per_second": 29.613,
|
| 827 |
+
"eval_steps_per_second": 3.736,
|
| 828 |
+
"step": 450
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.39535883111302106,
|
| 832 |
+
"grad_norm": 2.223949432373047,
|
| 833 |
+
"learning_rate": 3.92678391921108e-06,
|
| 834 |
+
"logits/chosen": 2.651564598083496,
|
| 835 |
+
"logits/rejected": 2.383842945098877,
|
| 836 |
+
"logps/chosen": -2.591308355331421,
|
| 837 |
+
"logps/rejected": -5.308972358703613,
|
| 838 |
+
"loss": 0.3412,
|
| 839 |
+
"rewards/accuracies": 0.762499988079071,
|
| 840 |
+
"rewards/chosen": -3.886962413787842,
|
| 841 |
+
"rewards/margins": 4.07649564743042,
|
| 842 |
+
"rewards/rejected": -7.963458061218262,
|
| 843 |
+
"step": 460
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.4039535883111302,
|
| 847 |
+
"grad_norm": 3.110624074935913,
|
| 848 |
+
"learning_rate": 3.88347887310836e-06,
|
| 849 |
+
"logits/chosen": 2.5435309410095215,
|
| 850 |
+
"logits/rejected": 2.46763277053833,
|
| 851 |
+
"logps/chosen": -2.413583993911743,
|
| 852 |
+
"logps/rejected": -5.543262481689453,
|
| 853 |
+
"loss": 0.3832,
|
| 854 |
+
"rewards/accuracies": 0.824999988079071,
|
| 855 |
+
"rewards/chosen": -3.620375871658325,
|
| 856 |
+
"rewards/margins": 4.694517135620117,
|
| 857 |
+
"rewards/rejected": -8.314892768859863,
|
| 858 |
+
"step": 470
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.41254834550923936,
|
| 862 |
+
"grad_norm": 1.6255794763565063,
|
| 863 |
+
"learning_rate": 3.839566987447492e-06,
|
| 864 |
+
"logits/chosen": 3.842928409576416,
|
| 865 |
+
"logits/rejected": 3.5797982215881348,
|
| 866 |
+
"logps/chosen": -2.6448044776916504,
|
| 867 |
+
"logps/rejected": -4.98160982131958,
|
| 868 |
+
"loss": 0.3547,
|
| 869 |
+
"rewards/accuracies": 0.8125,
|
| 870 |
+
"rewards/chosen": -3.9672069549560547,
|
| 871 |
+
"rewards/margins": 3.5052082538604736,
|
| 872 |
+
"rewards/rejected": -7.472414493560791,
|
| 873 |
+
"step": 480
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.4211431027073485,
|
| 877 |
+
"grad_norm": 2.9274284839630127,
|
| 878 |
+
"learning_rate": 3.795067523432826e-06,
|
| 879 |
+
"logits/chosen": 3.3297150135040283,
|
| 880 |
+
"logits/rejected": 3.0205535888671875,
|
| 881 |
+
"logps/chosen": -2.811923027038574,
|
| 882 |
+
"logps/rejected": -6.040881156921387,
|
| 883 |
+
"loss": 0.3097,
|
| 884 |
+
"rewards/accuracies": 0.887499988079071,
|
| 885 |
+
"rewards/chosen": -4.217884063720703,
|
| 886 |
+
"rewards/margins": 4.843437194824219,
|
| 887 |
+
"rewards/rejected": -9.061322212219238,
|
| 888 |
+
"step": 490
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 0.42973785990545765,
|
| 892 |
+
"grad_norm": 2.9143636226654053,
|
| 893 |
+
"learning_rate": 3.7500000000000005e-06,
|
| 894 |
+
"logits/chosen": 2.760014772415161,
|
| 895 |
+
"logits/rejected": 2.535520315170288,
|
| 896 |
+
"logps/chosen": -3.068406820297241,
|
| 897 |
+
"logps/rejected": -5.877435684204102,
|
| 898 |
+
"loss": 0.3031,
|
| 899 |
+
"rewards/accuracies": 0.875,
|
| 900 |
+
"rewards/chosen": -4.602609634399414,
|
| 901 |
+
"rewards/margins": 4.21354341506958,
|
| 902 |
+
"rewards/rejected": -8.816153526306152,
|
| 903 |
+
"step": 500
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.42973785990545765,
|
| 907 |
+
"eval_logits/chosen": 2.0952131748199463,
|
| 908 |
+
"eval_logits/rejected": 2.1864659786224365,
|
| 909 |
+
"eval_logps/chosen": -3.392296075820923,
|
| 910 |
+
"eval_logps/rejected": -6.948195457458496,
|
| 911 |
+
"eval_loss": 0.33660775423049927,
|
| 912 |
+
"eval_rewards/accuracies": 0.9263157844543457,
|
| 913 |
+
"eval_rewards/chosen": -5.088444232940674,
|
| 914 |
+
"eval_rewards/margins": 5.3338494300842285,
|
| 915 |
+
"eval_rewards/rejected": -10.422293663024902,
|
| 916 |
+
"eval_runtime": 25.4226,
|
| 917 |
+
"eval_samples_per_second": 29.619,
|
| 918 |
+
"eval_steps_per_second": 3.737,
|
| 919 |
+
"step": 500
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.4383326171035668,
|
| 923 |
+
"grad_norm": 2.563810348510742,
|
| 924 |
+
"learning_rate": 3.7043841852542884e-06,
|
| 925 |
+
"logits/chosen": 2.950286388397217,
|
| 926 |
+
"logits/rejected": 2.619025945663452,
|
| 927 |
+
"logps/chosen": -3.237391710281372,
|
| 928 |
+
"logps/rejected": -5.953216552734375,
|
| 929 |
+
"loss": 0.318,
|
| 930 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 931 |
+
"rewards/chosen": -4.856087684631348,
|
| 932 |
+
"rewards/margins": 4.073737144470215,
|
| 933 |
+
"rewards/rejected": -8.929824829101562,
|
| 934 |
+
"step": 510
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.44692737430167595,
|
| 938 |
+
"grad_norm": 2.0339434146881104,
|
| 939 |
+
"learning_rate": 3.658240087799655e-06,
|
| 940 |
+
"logits/chosen": 2.987595558166504,
|
| 941 |
+
"logits/rejected": 2.6243975162506104,
|
| 942 |
+
"logps/chosen": -3.5633530616760254,
|
| 943 |
+
"logps/rejected": -7.0458879470825195,
|
| 944 |
+
"loss": 0.3053,
|
| 945 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 946 |
+
"rewards/chosen": -5.345029354095459,
|
| 947 |
+
"rewards/margins": 5.223802089691162,
|
| 948 |
+
"rewards/rejected": -10.568831443786621,
|
| 949 |
+
"step": 520
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 0.45552213149978515,
|
| 953 |
+
"grad_norm": 4.091029644012451,
|
| 954 |
+
"learning_rate": 3.611587947962319e-06,
|
| 955 |
+
"logits/chosen": 2.297576904296875,
|
| 956 |
+
"logits/rejected": 2.0218777656555176,
|
| 957 |
+
"logps/chosen": -3.297245502471924,
|
| 958 |
+
"logps/rejected": -6.101919651031494,
|
| 959 |
+
"loss": 0.3255,
|
| 960 |
+
"rewards/accuracies": 0.887499988079071,
|
| 961 |
+
"rewards/chosen": -4.945868015289307,
|
| 962 |
+
"rewards/margins": 4.207010746002197,
|
| 963 |
+
"rewards/rejected": -9.152878761291504,
|
| 964 |
+
"step": 530
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.4641168886978943,
|
| 968 |
+
"grad_norm": 2.7896900177001953,
|
| 969 |
+
"learning_rate": 3.564448228912682e-06,
|
| 970 |
+
"logits/chosen": 2.103950023651123,
|
| 971 |
+
"logits/rejected": 1.9478647708892822,
|
| 972 |
+
"logps/chosen": -2.9360263347625732,
|
| 973 |
+
"logps/rejected": -6.406435489654541,
|
| 974 |
+
"loss": 0.3361,
|
| 975 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 976 |
+
"rewards/chosen": -4.40403938293457,
|
| 977 |
+
"rewards/margins": 5.20561408996582,
|
| 978 |
+
"rewards/rejected": -9.60965347290039,
|
| 979 |
+
"step": 540
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 0.47271164589600345,
|
| 983 |
+
"grad_norm": 2.657970905303955,
|
| 984 |
+
"learning_rate": 3.516841607689501e-06,
|
| 985 |
+
"logits/chosen": 2.1658639907836914,
|
| 986 |
+
"logits/rejected": 2.214900493621826,
|
| 987 |
+
"logps/chosen": -3.084073066711426,
|
| 988 |
+
"logps/rejected": -6.935500144958496,
|
| 989 |
+
"loss": 0.2928,
|
| 990 |
+
"rewards/accuracies": 0.862500011920929,
|
| 991 |
+
"rewards/chosen": -4.626110076904297,
|
| 992 |
+
"rewards/margins": 5.7771406173706055,
|
| 993 |
+
"rewards/rejected": -10.403249740600586,
|
| 994 |
+
"step": 550
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.47271164589600345,
|
| 998 |
+
"eval_logits/chosen": 2.285294771194458,
|
| 999 |
+
"eval_logits/rejected": 2.3312103748321533,
|
| 1000 |
+
"eval_logps/chosen": -3.35794997215271,
|
| 1001 |
+
"eval_logps/rejected": -7.37537145614624,
|
| 1002 |
+
"eval_loss": 0.3121817409992218,
|
| 1003 |
+
"eval_rewards/accuracies": 0.9263157844543457,
|
| 1004 |
+
"eval_rewards/chosen": -5.036925792694092,
|
| 1005 |
+
"eval_rewards/margins": 6.026132106781006,
|
| 1006 |
+
"eval_rewards/rejected": -11.063057899475098,
|
| 1007 |
+
"eval_runtime": 25.4015,
|
| 1008 |
+
"eval_samples_per_second": 29.644,
|
| 1009 |
+
"eval_steps_per_second": 3.74,
|
| 1010 |
+
"step": 550
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.4813064030941126,
|
| 1014 |
+
"grad_norm": 2.940019369125366,
|
| 1015 |
+
"learning_rate": 3.4687889661302577e-06,
|
| 1016 |
+
"logits/chosen": 1.9122416973114014,
|
| 1017 |
+
"logits/rejected": 1.9943454265594482,
|
| 1018 |
+
"logps/chosen": -3.27177095413208,
|
| 1019 |
+
"logps/rejected": -7.023342132568359,
|
| 1020 |
+
"loss": 0.3105,
|
| 1021 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1022 |
+
"rewards/chosen": -4.907656669616699,
|
| 1023 |
+
"rewards/margins": 5.6273579597473145,
|
| 1024 |
+
"rewards/rejected": -10.535014152526855,
|
| 1025 |
+
"step": 560
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.48990116029222175,
|
| 1029 |
+
"grad_norm": 1.8887412548065186,
|
| 1030 |
+
"learning_rate": 3.4203113817116955e-06,
|
| 1031 |
+
"logits/chosen": 2.274843692779541,
|
| 1032 |
+
"logits/rejected": 2.392199993133545,
|
| 1033 |
+
"logps/chosen": -3.383749008178711,
|
| 1034 |
+
"logps/rejected": -7.265415191650391,
|
| 1035 |
+
"loss": 0.3003,
|
| 1036 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1037 |
+
"rewards/chosen": -5.075623512268066,
|
| 1038 |
+
"rewards/margins": 5.8224992752075195,
|
| 1039 |
+
"rewards/rejected": -10.898123741149902,
|
| 1040 |
+
"step": 570
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.4984959174903309,
|
| 1044 |
+
"grad_norm": 1.6364414691925049,
|
| 1045 |
+
"learning_rate": 3.3714301183045382e-06,
|
| 1046 |
+
"logits/chosen": 2.423910617828369,
|
| 1047 |
+
"logits/rejected": 2.244985818862915,
|
| 1048 |
+
"logps/chosen": -3.0959205627441406,
|
| 1049 |
+
"logps/rejected": -6.822405815124512,
|
| 1050 |
+
"loss": 0.2471,
|
| 1051 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1052 |
+
"rewards/chosen": -4.643880844116211,
|
| 1053 |
+
"rewards/margins": 5.58972692489624,
|
| 1054 |
+
"rewards/rejected": -10.233609199523926,
|
| 1055 |
+
"step": 580
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 0.50709067468844,
|
| 1059 |
+
"grad_norm": 2.6540188789367676,
|
| 1060 |
+
"learning_rate": 3.3221666168464584e-06,
|
| 1061 |
+
"logits/chosen": 2.8146812915802,
|
| 1062 |
+
"logits/rejected": 2.5971922874450684,
|
| 1063 |
+
"logps/chosen": -4.139407157897949,
|
| 1064 |
+
"logps/rejected": -7.71649694442749,
|
| 1065 |
+
"loss": 0.2809,
|
| 1066 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1067 |
+
"rewards/chosen": -6.209111213684082,
|
| 1068 |
+
"rewards/margins": 5.365634441375732,
|
| 1069 |
+
"rewards/rejected": -11.574746131896973,
|
| 1070 |
+
"step": 590
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.5156854318865493,
|
| 1074 |
+
"grad_norm": 4.229885578155518,
|
| 1075 |
+
"learning_rate": 3.272542485937369e-06,
|
| 1076 |
+
"logits/chosen": 2.2735249996185303,
|
| 1077 |
+
"logits/rejected": 1.8577899932861328,
|
| 1078 |
+
"logps/chosen": -3.731342315673828,
|
| 1079 |
+
"logps/rejected": -7.2900390625,
|
| 1080 |
+
"loss": 0.2956,
|
| 1081 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1082 |
+
"rewards/chosen": -5.5970139503479,
|
| 1083 |
+
"rewards/margins": 5.338044166564941,
|
| 1084 |
+
"rewards/rejected": -10.93505859375,
|
| 1085 |
+
"step": 600
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 0.5156854318865493,
|
| 1089 |
+
"eval_logits/chosen": 2.3333992958068848,
|
| 1090 |
+
"eval_logits/rejected": 2.529745578765869,
|
| 1091 |
+
"eval_logps/chosen": -3.679597854614258,
|
| 1092 |
+
"eval_logps/rejected": -7.917842864990234,
|
| 1093 |
+
"eval_loss": 0.3030374050140381,
|
| 1094 |
+
"eval_rewards/accuracies": 0.9263157844543457,
|
| 1095 |
+
"eval_rewards/chosen": -5.519396781921387,
|
| 1096 |
+
"eval_rewards/margins": 6.357367992401123,
|
| 1097 |
+
"eval_rewards/rejected": -11.876765251159668,
|
| 1098 |
+
"eval_runtime": 25.5622,
|
| 1099 |
+
"eval_samples_per_second": 29.458,
|
| 1100 |
+
"eval_steps_per_second": 3.716,
|
| 1101 |
+
"step": 600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.5242801890846583,
|
| 1105 |
+
"grad_norm": 2.657008647918701,
|
| 1106 |
+
"learning_rate": 3.222579492361179e-06,
|
| 1107 |
+
"logits/chosen": 2.699007034301758,
|
| 1108 |
+
"logits/rejected": 2.731860876083374,
|
| 1109 |
+
"logps/chosen": -3.3311946392059326,
|
| 1110 |
+
"logps/rejected": -7.005735874176025,
|
| 1111 |
+
"loss": 0.2898,
|
| 1112 |
+
"rewards/accuracies": 0.9375,
|
| 1113 |
+
"rewards/chosen": -4.996791839599609,
|
| 1114 |
+
"rewards/margins": 5.511812686920166,
|
| 1115 |
+
"rewards/rejected": -10.508604049682617,
|
| 1116 |
+
"step": 610
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.5328749462827675,
|
| 1120 |
+
"grad_norm": 3.046638250350952,
|
| 1121 |
+
"learning_rate": 3.1722995515381644e-06,
|
| 1122 |
+
"logits/chosen": 2.7617671489715576,
|
| 1123 |
+
"logits/rejected": 2.7338194847106934,
|
| 1124 |
+
"logps/chosen": -3.336381435394287,
|
| 1125 |
+
"logps/rejected": -7.058961391448975,
|
| 1126 |
+
"loss": 0.2895,
|
| 1127 |
+
"rewards/accuracies": 0.9375,
|
| 1128 |
+
"rewards/chosen": -5.004572868347168,
|
| 1129 |
+
"rewards/margins": 5.583868980407715,
|
| 1130 |
+
"rewards/rejected": -10.588441848754883,
|
| 1131 |
+
"step": 620
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.5414697034808766,
|
| 1135 |
+
"grad_norm": 2.342069387435913,
|
| 1136 |
+
"learning_rate": 3.121724717912138e-06,
|
| 1137 |
+
"logits/chosen": 2.5818216800689697,
|
| 1138 |
+
"logits/rejected": 1.987378716468811,
|
| 1139 |
+
"logps/chosen": -3.0970518589019775,
|
| 1140 |
+
"logps/rejected": -6.240235805511475,
|
| 1141 |
+
"loss": 0.2634,
|
| 1142 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1143 |
+
"rewards/chosen": -4.645577430725098,
|
| 1144 |
+
"rewards/margins": 4.714776039123535,
|
| 1145 |
+
"rewards/rejected": -9.36035442352295,
|
| 1146 |
+
"step": 630
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.5500644606789858,
|
| 1150 |
+
"grad_norm": 1.9333513975143433,
|
| 1151 |
+
"learning_rate": 3.0708771752766397e-06,
|
| 1152 |
+
"logits/chosen": 2.911674737930298,
|
| 1153 |
+
"logits/rejected": 2.7606472969055176,
|
| 1154 |
+
"logps/chosen": -3.2809441089630127,
|
| 1155 |
+
"logps/rejected": -7.210829257965088,
|
| 1156 |
+
"loss": 0.2594,
|
| 1157 |
+
"rewards/accuracies": 0.9375,
|
| 1158 |
+
"rewards/chosen": -4.921416282653809,
|
| 1159 |
+
"rewards/margins": 5.894827365875244,
|
| 1160 |
+
"rewards/rejected": -10.816244125366211,
|
| 1161 |
+
"step": 640
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.5586592178770949,
|
| 1165 |
+
"grad_norm": 5.659445285797119,
|
| 1166 |
+
"learning_rate": 3.019779227044398e-06,
|
| 1167 |
+
"logits/chosen": 2.4733409881591797,
|
| 1168 |
+
"logits/rejected": 2.102668285369873,
|
| 1169 |
+
"logps/chosen": -3.4448726177215576,
|
| 1170 |
+
"logps/rejected": -7.304962158203125,
|
| 1171 |
+
"loss": 0.2399,
|
| 1172 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1173 |
+
"rewards/chosen": -5.167309284210205,
|
| 1174 |
+
"rewards/margins": 5.790134429931641,
|
| 1175 |
+
"rewards/rejected": -10.957443237304688,
|
| 1176 |
+
"step": 650
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 0.5586592178770949,
|
| 1180 |
+
"eval_logits/chosen": 2.482032537460327,
|
| 1181 |
+
"eval_logits/rejected": 2.66147780418396,
|
| 1182 |
+
"eval_logps/chosen": -3.728013515472412,
|
| 1183 |
+
"eval_logps/rejected": -8.231985092163086,
|
| 1184 |
+
"eval_loss": 0.2814938426017761,
|
| 1185 |
+
"eval_rewards/accuracies": 0.9263157844543457,
|
| 1186 |
+
"eval_rewards/chosen": -5.592020511627197,
|
| 1187 |
+
"eval_rewards/margins": 6.75595760345459,
|
| 1188 |
+
"eval_rewards/rejected": -12.347977638244629,
|
| 1189 |
+
"eval_runtime": 25.4252,
|
| 1190 |
+
"eval_samples_per_second": 29.616,
|
| 1191 |
+
"eval_steps_per_second": 3.736,
|
| 1192 |
+
"step": 650
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.5672539750752041,
|
| 1196 |
+
"grad_norm": 2.189638137817383,
|
| 1197 |
+
"learning_rate": 2.9684532864643123e-06,
|
| 1198 |
+
"logits/chosen": 2.875077962875366,
|
| 1199 |
+
"logits/rejected": 2.712646484375,
|
| 1200 |
+
"logps/chosen": -3.757338762283325,
|
| 1201 |
+
"logps/rejected": -6.6974897384643555,
|
| 1202 |
+
"loss": 0.2759,
|
| 1203 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1204 |
+
"rewards/chosen": -5.636007785797119,
|
| 1205 |
+
"rewards/margins": 4.410226821899414,
|
| 1206 |
+
"rewards/rejected": -10.046236038208008,
|
| 1207 |
+
"step": 660
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.5758487322733132,
|
| 1211 |
+
"grad_norm": 3.5755774974823,
|
| 1212 |
+
"learning_rate": 2.9169218667902562e-06,
|
| 1213 |
+
"logits/chosen": 2.9562981128692627,
|
| 1214 |
+
"logits/rejected": 2.7660539150238037,
|
| 1215 |
+
"logps/chosen": -3.2358715534210205,
|
| 1216 |
+
"logps/rejected": -6.90399169921875,
|
| 1217 |
+
"loss": 0.2586,
|
| 1218 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1219 |
+
"rewards/chosen": -4.853806972503662,
|
| 1220 |
+
"rewards/margins": 5.502181053161621,
|
| 1221 |
+
"rewards/rejected": -10.355987548828125,
|
| 1222 |
+
"step": 670
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 0.5844434894714224,
|
| 1226 |
+
"grad_norm": 2.5616958141326904,
|
| 1227 |
+
"learning_rate": 2.8652075714060296e-06,
|
| 1228 |
+
"logits/chosen": 2.5067126750946045,
|
| 1229 |
+
"logits/rejected": 2.3888354301452637,
|
| 1230 |
+
"logps/chosen": -3.462563991546631,
|
| 1231 |
+
"logps/rejected": -6.964964866638184,
|
| 1232 |
+
"loss": 0.251,
|
| 1233 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1234 |
+
"rewards/chosen": -5.193846225738525,
|
| 1235 |
+
"rewards/margins": 5.253602027893066,
|
| 1236 |
+
"rewards/rejected": -10.447446823120117,
|
| 1237 |
+
"step": 680
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 0.5930382466695315,
|
| 1241 |
+
"grad_norm": 2.964050531387329,
|
| 1242 |
+
"learning_rate": 2.813333083910761e-06,
|
| 1243 |
+
"logits/chosen": 2.659935474395752,
|
| 1244 |
+
"logits/rejected": 2.6573758125305176,
|
| 1245 |
+
"logps/chosen": -3.9107768535614014,
|
| 1246 |
+
"logps/rejected": -7.865903377532959,
|
| 1247 |
+
"loss": 0.2294,
|
| 1248 |
+
"rewards/accuracies": 0.9375,
|
| 1249 |
+
"rewards/chosen": -5.866166114807129,
|
| 1250 |
+
"rewards/margins": 5.9326887130737305,
|
| 1251 |
+
"rewards/rejected": -11.79885482788086,
|
| 1252 |
+
"step": 690
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.6016330038676407,
|
| 1256 |
+
"grad_norm": 4.389697551727295,
|
| 1257 |
+
"learning_rate": 2.761321158169134e-06,
|
| 1258 |
+
"logits/chosen": 2.217245578765869,
|
| 1259 |
+
"logits/rejected": 2.421597957611084,
|
| 1260 |
+
"logps/chosen": -4.029661655426025,
|
| 1261 |
+
"logps/rejected": -8.073125839233398,
|
| 1262 |
+
"loss": 0.2469,
|
| 1263 |
+
"rewards/accuracies": 0.9375,
|
| 1264 |
+
"rewards/chosen": -6.044493675231934,
|
| 1265 |
+
"rewards/margins": 6.065195083618164,
|
| 1266 |
+
"rewards/rejected": -12.109688758850098,
|
| 1267 |
+
"step": 700
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 0.6016330038676407,
|
| 1271 |
+
"eval_logits/chosen": 2.0770955085754395,
|
| 1272 |
+
"eval_logits/rejected": 2.3815462589263916,
|
| 1273 |
+
"eval_logps/chosen": -3.924149751663208,
|
| 1274 |
+
"eval_logps/rejected": -8.844257354736328,
|
| 1275 |
+
"eval_loss": 0.2584603726863861,
|
| 1276 |
+
"eval_rewards/accuracies": 0.9263157844543457,
|
| 1277 |
+
"eval_rewards/chosen": -5.886224746704102,
|
| 1278 |
+
"eval_rewards/margins": 7.380159854888916,
|
| 1279 |
+
"eval_rewards/rejected": -13.26638412475586,
|
| 1280 |
+
"eval_runtime": 25.4228,
|
| 1281 |
+
"eval_samples_per_second": 29.619,
|
| 1282 |
+
"eval_steps_per_second": 3.737,
|
| 1283 |
+
"step": 700
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.6102277610657499,
|
| 1287 |
+
"grad_norm": 3.290154457092285,
|
| 1288 |
+
"learning_rate": 2.70919460833079e-06,
|
| 1289 |
+
"logits/chosen": 2.458578586578369,
|
| 1290 |
+
"logits/rejected": 2.275515079498291,
|
| 1291 |
+
"logps/chosen": -3.2734694480895996,
|
| 1292 |
+
"logps/rejected": -7.873226165771484,
|
| 1293 |
+
"loss": 0.2732,
|
| 1294 |
+
"rewards/accuracies": 0.9375,
|
| 1295 |
+
"rewards/chosen": -4.91020393371582,
|
| 1296 |
+
"rewards/margins": 6.899635314941406,
|
| 1297 |
+
"rewards/rejected": -11.809839248657227,
|
| 1298 |
+
"step": 710
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.618822518263859,
|
| 1302 |
+
"grad_norm": 2.2760908603668213,
|
| 1303 |
+
"learning_rate": 2.6569762988232838e-06,
|
| 1304 |
+
"logits/chosen": 2.6856372356414795,
|
| 1305 |
+
"logits/rejected": 2.722838878631592,
|
| 1306 |
+
"logps/chosen": -3.589418411254883,
|
| 1307 |
+
"logps/rejected": -7.638446807861328,
|
| 1308 |
+
"loss": 0.2583,
|
| 1309 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1310 |
+
"rewards/chosen": -5.384127140045166,
|
| 1311 |
+
"rewards/margins": 6.073542594909668,
|
| 1312 |
+
"rewards/rejected": -11.457670211791992,
|
| 1313 |
+
"step": 720
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 0.6274172754619682,
|
| 1317 |
+
"grad_norm": 6.937672138214111,
|
| 1318 |
+
"learning_rate": 2.604689134322999e-06,
|
| 1319 |
+
"logits/chosen": 2.928969383239746,
|
| 1320 |
+
"logits/rejected": 2.5493836402893066,
|
| 1321 |
+
"logps/chosen": -3.3862743377685547,
|
| 1322 |
+
"logps/rejected": -7.568005561828613,
|
| 1323 |
+
"loss": 0.2889,
|
| 1324 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1325 |
+
"rewards/chosen": -5.07941198348999,
|
| 1326 |
+
"rewards/margins": 6.27259635925293,
|
| 1327 |
+
"rewards/rejected": -11.352007865905762,
|
| 1328 |
+
"step": 730
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.6360120326600773,
|
| 1332 |
+
"grad_norm": 2.1878838539123535,
|
| 1333 |
+
"learning_rate": 2.5523560497083927e-06,
|
| 1334 |
+
"logits/chosen": 2.3824827671051025,
|
| 1335 |
+
"logits/rejected": 2.257145404815674,
|
| 1336 |
+
"logps/chosen": -3.5448341369628906,
|
| 1337 |
+
"logps/rejected": -7.594444274902344,
|
| 1338 |
+
"loss": 0.1972,
|
| 1339 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1340 |
+
"rewards/chosen": -5.3172502517700195,
|
| 1341 |
+
"rewards/margins": 6.074415683746338,
|
| 1342 |
+
"rewards/rejected": -11.3916654586792,
|
| 1343 |
+
"step": 740
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.6446067898581865,
|
| 1347 |
+
"grad_norm": 4.405832767486572,
|
| 1348 |
+
"learning_rate": 2.5e-06,
|
| 1349 |
+
"logits/chosen": 3.204157590866089,
|
| 1350 |
+
"logits/rejected": 3.0262837409973145,
|
| 1351 |
+
"logps/chosen": -3.67409086227417,
|
| 1352 |
+
"logps/rejected": -8.078901290893555,
|
| 1353 |
+
"loss": 0.2282,
|
| 1354 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1355 |
+
"rewards/chosen": -5.511136531829834,
|
| 1356 |
+
"rewards/margins": 6.607214450836182,
|
| 1357 |
+
"rewards/rejected": -12.118351936340332,
|
| 1358 |
+
"step": 750
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 0.6446067898581865,
|
| 1362 |
+
"eval_logits/chosen": 2.1246254444122314,
|
| 1363 |
+
"eval_logits/rejected": 2.4088852405548096,
|
| 1364 |
+
"eval_logps/chosen": -4.221064567565918,
|
| 1365 |
+
"eval_logps/rejected": -9.4141206741333,
|
| 1366 |
+
"eval_loss": 0.2537557780742645,
|
| 1367 |
+
"eval_rewards/accuracies": 0.9368420839309692,
|
| 1368 |
+
"eval_rewards/chosen": -6.331596374511719,
|
| 1369 |
+
"eval_rewards/margins": 7.789584159851074,
|
| 1370 |
+
"eval_rewards/rejected": -14.121179580688477,
|
| 1371 |
+
"eval_runtime": 25.436,
|
| 1372 |
+
"eval_samples_per_second": 29.604,
|
| 1373 |
+
"eval_steps_per_second": 3.735,
|
| 1374 |
+
"step": 750
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.6532015470562956,
|
| 1378 |
+
"grad_norm": 2.8693907260894775,
|
| 1379 |
+
"learning_rate": 2.447643950291608e-06,
|
| 1380 |
+
"logits/chosen": 2.5033986568450928,
|
| 1381 |
+
"logits/rejected": 2.2746779918670654,
|
| 1382 |
+
"logps/chosen": -4.256644248962402,
|
| 1383 |
+
"logps/rejected": -8.564817428588867,
|
| 1384 |
+
"loss": 0.2337,
|
| 1385 |
+
"rewards/accuracies": 0.9375,
|
| 1386 |
+
"rewards/chosen": -6.3849663734436035,
|
| 1387 |
+
"rewards/margins": 6.462259769439697,
|
| 1388 |
+
"rewards/rejected": -12.8472261428833,
|
| 1389 |
+
"step": 760
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.6617963042544048,
|
| 1393 |
+
"grad_norm": 4.912906646728516,
|
| 1394 |
+
"learning_rate": 2.3953108656770018e-06,
|
| 1395 |
+
"logits/chosen": 2.861431837081909,
|
| 1396 |
+
"logits/rejected": 2.974611759185791,
|
| 1397 |
+
"logps/chosen": -3.9564735889434814,
|
| 1398 |
+
"logps/rejected": -7.863286018371582,
|
| 1399 |
+
"loss": 0.2585,
|
| 1400 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1401 |
+
"rewards/chosen": -5.934710502624512,
|
| 1402 |
+
"rewards/margins": 5.860217571258545,
|
| 1403 |
+
"rewards/rejected": -11.794927597045898,
|
| 1404 |
+
"step": 770
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 0.6703910614525139,
|
| 1408 |
+
"grad_norm": 3.215716600418091,
|
| 1409 |
+
"learning_rate": 2.3430237011767166e-06,
|
| 1410 |
+
"logits/chosen": 1.9008615016937256,
|
| 1411 |
+
"logits/rejected": 1.9049352407455444,
|
| 1412 |
+
"logps/chosen": -4.304060935974121,
|
| 1413 |
+
"logps/rejected": -8.806629180908203,
|
| 1414 |
+
"loss": 0.2279,
|
| 1415 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1416 |
+
"rewards/chosen": -6.456091403961182,
|
| 1417 |
+
"rewards/margins": 6.753852844238281,
|
| 1418 |
+
"rewards/rejected": -13.209943771362305,
|
| 1419 |
+
"step": 780
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.6789858186506231,
|
| 1423 |
+
"grad_norm": 3.8724021911621094,
|
| 1424 |
+
"learning_rate": 2.290805391669212e-06,
|
| 1425 |
+
"logits/chosen": 2.2521636486053467,
|
| 1426 |
+
"logits/rejected": 2.2159788608551025,
|
| 1427 |
+
"logps/chosen": -4.012774467468262,
|
| 1428 |
+
"logps/rejected": -8.53366470336914,
|
| 1429 |
+
"loss": 0.2437,
|
| 1430 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1431 |
+
"rewards/chosen": -6.019161701202393,
|
| 1432 |
+
"rewards/margins": 6.78133487701416,
|
| 1433 |
+
"rewards/rejected": -12.800497055053711,
|
| 1434 |
+
"step": 790
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.6875805758487322,
|
| 1438 |
+
"grad_norm": 3.56345796585083,
|
| 1439 |
+
"learning_rate": 2.238678841830867e-06,
|
| 1440 |
+
"logits/chosen": 2.0579304695129395,
|
| 1441 |
+
"logits/rejected": 2.304316997528076,
|
| 1442 |
+
"logps/chosen": -3.590430736541748,
|
| 1443 |
+
"logps/rejected": -8.182169914245605,
|
| 1444 |
+
"loss": 0.213,
|
| 1445 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1446 |
+
"rewards/chosen": -5.385646820068359,
|
| 1447 |
+
"rewards/margins": 6.887608528137207,
|
| 1448 |
+
"rewards/rejected": -12.27325439453125,
|
| 1449 |
+
"step": 800
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.6875805758487322,
|
| 1453 |
+
"eval_logits/chosen": 2.228646755218506,
|
| 1454 |
+
"eval_logits/rejected": 2.444817543029785,
|
| 1455 |
+
"eval_logps/chosen": -3.8403449058532715,
|
| 1456 |
+
"eval_logps/rejected": -9.179658889770508,
|
| 1457 |
+
"eval_loss": 0.23895224928855896,
|
| 1458 |
+
"eval_rewards/accuracies": 0.9368420839309692,
|
| 1459 |
+
"eval_rewards/chosen": -5.76051664352417,
|
| 1460 |
+
"eval_rewards/margins": 8.00897216796875,
|
| 1461 |
+
"eval_rewards/rejected": -13.769490242004395,
|
| 1462 |
+
"eval_runtime": 25.3925,
|
| 1463 |
+
"eval_samples_per_second": 29.654,
|
| 1464 |
+
"eval_steps_per_second": 3.741,
|
| 1465 |
+
"step": 800
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.6961753330468414,
|
| 1469 |
+
"grad_norm": 3.4880526065826416,
|
| 1470 |
+
"learning_rate": 2.186666916089239e-06,
|
| 1471 |
+
"logits/chosen": 1.7993383407592773,
|
| 1472 |
+
"logits/rejected": 1.754417061805725,
|
| 1473 |
+
"logps/chosen": -4.045234680175781,
|
| 1474 |
+
"logps/rejected": -8.927519798278809,
|
| 1475 |
+
"loss": 0.2391,
|
| 1476 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1477 |
+
"rewards/chosen": -6.067852020263672,
|
| 1478 |
+
"rewards/margins": 7.323427677154541,
|
| 1479 |
+
"rewards/rejected": -13.391279220581055,
|
| 1480 |
+
"step": 810
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.7047700902449506,
|
| 1484 |
+
"grad_norm": 3.56809139251709,
|
| 1485 |
+
"learning_rate": 2.134792428593971e-06,
|
| 1486 |
+
"logits/chosen": 2.9591994285583496,
|
| 1487 |
+
"logits/rejected": 2.960444211959839,
|
| 1488 |
+
"logps/chosen": -4.150156497955322,
|
| 1489 |
+
"logps/rejected": -8.512441635131836,
|
| 1490 |
+
"loss": 0.1972,
|
| 1491 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1492 |
+
"rewards/chosen": -6.225234031677246,
|
| 1493 |
+
"rewards/margins": 6.54342794418335,
|
| 1494 |
+
"rewards/rejected": -12.768662452697754,
|
| 1495 |
+
"step": 820
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 0.7133648474430597,
|
| 1499 |
+
"grad_norm": 4.127833843231201,
|
| 1500 |
+
"learning_rate": 2.0830781332097446e-06,
|
| 1501 |
+
"logits/chosen": 3.008269786834717,
|
| 1502 |
+
"logits/rejected": 2.63409686088562,
|
| 1503 |
+
"logps/chosen": -3.8291163444519043,
|
| 1504 |
+
"logps/rejected": -8.657347679138184,
|
| 1505 |
+
"loss": 0.2161,
|
| 1506 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1507 |
+
"rewards/chosen": -5.743674278259277,
|
| 1508 |
+
"rewards/margins": 7.24234676361084,
|
| 1509 |
+
"rewards/rejected": -12.986021041870117,
|
| 1510 |
+
"step": 830
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 0.7219596046411689,
|
| 1514 |
+
"grad_norm": 4.475767612457275,
|
| 1515 |
+
"learning_rate": 2.031546713535688e-06,
|
| 1516 |
+
"logits/chosen": 2.7164976596832275,
|
| 1517 |
+
"logits/rejected": 2.5976195335388184,
|
| 1518 |
+
"logps/chosen": -4.153134346008301,
|
| 1519 |
+
"logps/rejected": -8.893486022949219,
|
| 1520 |
+
"loss": 0.1895,
|
| 1521 |
+
"rewards/accuracies": 0.9375,
|
| 1522 |
+
"rewards/chosen": -6.229701519012451,
|
| 1523 |
+
"rewards/margins": 7.110527992248535,
|
| 1524 |
+
"rewards/rejected": -13.340228080749512,
|
| 1525 |
+
"step": 840
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 0.730554361839278,
|
| 1529 |
+
"grad_norm": 4.190205097198486,
|
| 1530 |
+
"learning_rate": 1.9802207729556023e-06,
|
| 1531 |
+
"logits/chosen": 2.6235451698303223,
|
| 1532 |
+
"logits/rejected": 2.5486202239990234,
|
| 1533 |
+
"logps/chosen": -3.899543046951294,
|
| 1534 |
+
"logps/rejected": -8.277327537536621,
|
| 1535 |
+
"loss": 0.2239,
|
| 1536 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1537 |
+
"rewards/chosen": -5.8493146896362305,
|
| 1538 |
+
"rewards/margins": 6.566677093505859,
|
| 1539 |
+
"rewards/rejected": -12.415990829467773,
|
| 1540 |
+
"step": 850
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 0.730554361839278,
|
| 1544 |
+
"eval_logits/chosen": 2.173233985900879,
|
| 1545 |
+
"eval_logits/rejected": 2.433162212371826,
|
| 1546 |
+
"eval_logps/chosen": -4.13487434387207,
|
| 1547 |
+
"eval_logps/rejected": -9.577596664428711,
|
| 1548 |
+
"eval_loss": 0.23591776192188263,
|
| 1549 |
+
"eval_rewards/accuracies": 0.9473684430122375,
|
| 1550 |
+
"eval_rewards/chosen": -6.2023115158081055,
|
| 1551 |
+
"eval_rewards/margins": 8.164085388183594,
|
| 1552 |
+
"eval_rewards/rejected": -14.3663969039917,
|
| 1553 |
+
"eval_runtime": 25.4513,
|
| 1554 |
+
"eval_samples_per_second": 29.586,
|
| 1555 |
+
"eval_steps_per_second": 3.733,
|
| 1556 |
+
"step": 850
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.7391491190373872,
|
| 1560 |
+
"grad_norm": 2.6548664569854736,
|
| 1561 |
+
"learning_rate": 1.9291228247233607e-06,
|
| 1562 |
+
"logits/chosen": 1.7737414836883545,
|
| 1563 |
+
"logits/rejected": 2.080662965774536,
|
| 1564 |
+
"logps/chosen": -3.9447720050811768,
|
| 1565 |
+
"logps/rejected": -9.01865005493164,
|
| 1566 |
+
"loss": 0.2268,
|
| 1567 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1568 |
+
"rewards/chosen": -5.9171576499938965,
|
| 1569 |
+
"rewards/margins": 7.610815525054932,
|
| 1570 |
+
"rewards/rejected": -13.527974128723145,
|
| 1571 |
+
"step": 860
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.7477438762354963,
|
| 1575 |
+
"grad_norm": 2.5912184715270996,
|
| 1576 |
+
"learning_rate": 1.8782752820878636e-06,
|
| 1577 |
+
"logits/chosen": 2.5428760051727295,
|
| 1578 |
+
"logits/rejected": 2.3569278717041016,
|
| 1579 |
+
"logps/chosen": -3.685049057006836,
|
| 1580 |
+
"logps/rejected": -9.194517135620117,
|
| 1581 |
+
"loss": 0.2001,
|
| 1582 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1583 |
+
"rewards/chosen": -5.527573585510254,
|
| 1584 |
+
"rewards/margins": 8.264203071594238,
|
| 1585 |
+
"rewards/rejected": -13.791775703430176,
|
| 1586 |
+
"step": 870
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.7563386334336055,
|
| 1590 |
+
"grad_norm": 3.789594888687134,
|
| 1591 |
+
"learning_rate": 1.827700448461836e-06,
|
| 1592 |
+
"logits/chosen": 3.139338970184326,
|
| 1593 |
+
"logits/rejected": 3.003114700317383,
|
| 1594 |
+
"logps/chosen": -4.347461700439453,
|
| 1595 |
+
"logps/rejected": -8.560078620910645,
|
| 1596 |
+
"loss": 0.2257,
|
| 1597 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1598 |
+
"rewards/chosen": -6.521193027496338,
|
| 1599 |
+
"rewards/margins": 6.318924903869629,
|
| 1600 |
+
"rewards/rejected": -12.840118408203125,
|
| 1601 |
+
"step": 880
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.7649333906317146,
|
| 1605 |
+
"grad_norm": 2.3799326419830322,
|
| 1606 |
+
"learning_rate": 1.7774205076388207e-06,
|
| 1607 |
+
"logits/chosen": 3.2622504234313965,
|
| 1608 |
+
"logits/rejected": 2.922945261001587,
|
| 1609 |
+
"logps/chosen": -4.306991100311279,
|
| 1610 |
+
"logps/rejected": -8.622769355773926,
|
| 1611 |
+
"loss": 0.2123,
|
| 1612 |
+
"rewards/accuracies": 0.9375,
|
| 1613 |
+
"rewards/chosen": -6.46048641204834,
|
| 1614 |
+
"rewards/margins": 6.473666191101074,
|
| 1615 |
+
"rewards/rejected": -12.93415355682373,
|
| 1616 |
+
"step": 890
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.7735281478298238,
|
| 1620 |
+
"grad_norm": 3.4133400917053223,
|
| 1621 |
+
"learning_rate": 1.7274575140626318e-06,
|
| 1622 |
+
"logits/chosen": 2.8558292388916016,
|
| 1623 |
+
"logits/rejected": 2.919982433319092,
|
| 1624 |
+
"logps/chosen": -3.791405200958252,
|
| 1625 |
+
"logps/rejected": -9.348276138305664,
|
| 1626 |
+
"loss": 0.2345,
|
| 1627 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1628 |
+
"rewards/chosen": -5.687107086181641,
|
| 1629 |
+
"rewards/margins": 8.335307121276855,
|
| 1630 |
+
"rewards/rejected": -14.022415161132812,
|
| 1631 |
+
"step": 900
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 0.7735281478298238,
|
| 1635 |
+
"eval_logits/chosen": 2.2851152420043945,
|
| 1636 |
+
"eval_logits/rejected": 2.5511629581451416,
|
| 1637 |
+
"eval_logps/chosen": -4.023584842681885,
|
| 1638 |
+
"eval_logps/rejected": -9.625852584838867,
|
| 1639 |
+
"eval_loss": 0.23031750321388245,
|
| 1640 |
+
"eval_rewards/accuracies": 0.9473684430122375,
|
| 1641 |
+
"eval_rewards/chosen": -6.035377025604248,
|
| 1642 |
+
"eval_rewards/margins": 8.403401374816895,
|
| 1643 |
+
"eval_rewards/rejected": -14.4387788772583,
|
| 1644 |
+
"eval_runtime": 25.4061,
|
| 1645 |
+
"eval_samples_per_second": 29.639,
|
| 1646 |
+
"eval_steps_per_second": 3.739,
|
| 1647 |
+
"step": 900
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.7821229050279329,
|
| 1651 |
+
"grad_norm": 2.178900957107544,
|
| 1652 |
+
"learning_rate": 1.677833383153542e-06,
|
| 1653 |
+
"logits/chosen": 2.3001868724823,
|
| 1654 |
+
"logits/rejected": 2.365304470062256,
|
| 1655 |
+
"logps/chosen": -3.690169095993042,
|
| 1656 |
+
"logps/rejected": -8.727324485778809,
|
| 1657 |
+
"loss": 0.1988,
|
| 1658 |
+
"rewards/accuracies": 0.9375,
|
| 1659 |
+
"rewards/chosen": -5.535253047943115,
|
| 1660 |
+
"rewards/margins": 7.555734157562256,
|
| 1661 |
+
"rewards/rejected": -13.090988159179688,
|
| 1662 |
+
"step": 910
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.7907176622260421,
|
| 1666 |
+
"grad_norm": 4.60929536819458,
|
| 1667 |
+
"learning_rate": 1.6285698816954626e-06,
|
| 1668 |
+
"logits/chosen": 3.103785276412964,
|
| 1669 |
+
"logits/rejected": 3.0096678733825684,
|
| 1670 |
+
"logps/chosen": -4.184874534606934,
|
| 1671 |
+
"logps/rejected": -8.704519271850586,
|
| 1672 |
+
"loss": 0.2128,
|
| 1673 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1674 |
+
"rewards/chosen": -6.277312278747559,
|
| 1675 |
+
"rewards/margins": 6.7794671058654785,
|
| 1676 |
+
"rewards/rejected": -13.056779861450195,
|
| 1677 |
+
"step": 920
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.7993124194241513,
|
| 1681 |
+
"grad_norm": 1.1031241416931152,
|
| 1682 |
+
"learning_rate": 1.5796886182883053e-06,
|
| 1683 |
+
"logits/chosen": 3.2616991996765137,
|
| 1684 |
+
"logits/rejected": 2.990100622177124,
|
| 1685 |
+
"logps/chosen": -4.041825771331787,
|
| 1686 |
+
"logps/rejected": -9.399754524230957,
|
| 1687 |
+
"loss": 0.2131,
|
| 1688 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1689 |
+
"rewards/chosen": -6.062739372253418,
|
| 1690 |
+
"rewards/margins": 8.03689193725586,
|
| 1691 |
+
"rewards/rejected": -14.099630355834961,
|
| 1692 |
+
"step": 930
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 0.8079071766222604,
|
| 1696 |
+
"grad_norm": 1.8013640642166138,
|
| 1697 |
+
"learning_rate": 1.5312110338697427e-06,
|
| 1698 |
+
"logits/chosen": 2.2281856536865234,
|
| 1699 |
+
"logits/rejected": 2.1705000400543213,
|
| 1700 |
+
"logps/chosen": -3.802743434906006,
|
| 1701 |
+
"logps/rejected": -8.745875358581543,
|
| 1702 |
+
"loss": 0.2211,
|
| 1703 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1704 |
+
"rewards/chosen": -5.704115867614746,
|
| 1705 |
+
"rewards/margins": 7.414697170257568,
|
| 1706 |
+
"rewards/rejected": -13.118814468383789,
|
| 1707 |
+
"step": 940
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 0.8165019338203696,
|
| 1711 |
+
"grad_norm": 5.369480609893799,
|
| 1712 |
+
"learning_rate": 1.4831583923105e-06,
|
| 1713 |
+
"logits/chosen": 2.0146822929382324,
|
| 1714 |
+
"logits/rejected": 2.0050222873687744,
|
| 1715 |
+
"logps/chosen": -4.013974189758301,
|
| 1716 |
+
"logps/rejected": -9.14311408996582,
|
| 1717 |
+
"loss": 0.2359,
|
| 1718 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1719 |
+
"rewards/chosen": -6.020960807800293,
|
| 1720 |
+
"rewards/margins": 7.693708896636963,
|
| 1721 |
+
"rewards/rejected": -13.71467113494873,
|
| 1722 |
+
"step": 950
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 0.8165019338203696,
|
| 1726 |
+
"eval_logits/chosen": 2.4391298294067383,
|
| 1727 |
+
"eval_logits/rejected": 2.693408250808716,
|
| 1728 |
+
"eval_logps/chosen": -3.877185106277466,
|
| 1729 |
+
"eval_logps/rejected": -9.608528137207031,
|
| 1730 |
+
"eval_loss": 0.2290637195110321,
|
| 1731 |
+
"eval_rewards/accuracies": 0.9473684430122375,
|
| 1732 |
+
"eval_rewards/chosen": -5.81577730178833,
|
| 1733 |
+
"eval_rewards/margins": 8.597016334533691,
|
| 1734 |
+
"eval_rewards/rejected": -14.41279411315918,
|
| 1735 |
+
"eval_runtime": 25.4752,
|
| 1736 |
+
"eval_samples_per_second": 29.558,
|
| 1737 |
+
"eval_steps_per_second": 3.729,
|
| 1738 |
+
"step": 950
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.8250966910184787,
|
| 1742 |
+
"grad_norm": 4.45559549331665,
|
| 1743 |
+
"learning_rate": 1.4355517710873184e-06,
|
| 1744 |
+
"logits/chosen": 3.141005516052246,
|
| 1745 |
+
"logits/rejected": 3.037994861602783,
|
| 1746 |
+
"logps/chosen": -4.300066947937012,
|
| 1747 |
+
"logps/rejected": -9.016799926757812,
|
| 1748 |
+
"loss": 0.2369,
|
| 1749 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1750 |
+
"rewards/chosen": -6.450100898742676,
|
| 1751 |
+
"rewards/margins": 7.075097560882568,
|
| 1752 |
+
"rewards/rejected": -13.525197982788086,
|
| 1753 |
+
"step": 960
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.8336914482165879,
|
| 1757 |
+
"grad_norm": 4.787370681762695,
|
| 1758 |
+
"learning_rate": 1.388412052037682e-06,
|
| 1759 |
+
"logits/chosen": 2.0776166915893555,
|
| 1760 |
+
"logits/rejected": 2.1846489906311035,
|
| 1761 |
+
"logps/chosen": -3.574514865875244,
|
| 1762 |
+
"logps/rejected": -9.252939224243164,
|
| 1763 |
+
"loss": 0.2339,
|
| 1764 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1765 |
+
"rewards/chosen": -5.3617730140686035,
|
| 1766 |
+
"rewards/margins": 8.517634391784668,
|
| 1767 |
+
"rewards/rejected": -13.879406929016113,
|
| 1768 |
+
"step": 970
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.842286205414697,
|
| 1772 |
+
"grad_norm": 2.436915397644043,
|
| 1773 |
+
"learning_rate": 1.3417599122003464e-06,
|
| 1774 |
+
"logits/chosen": 2.4205520153045654,
|
| 1775 |
+
"logits/rejected": 2.362619161605835,
|
| 1776 |
+
"logps/chosen": -3.942296266555786,
|
| 1777 |
+
"logps/rejected": -9.501001358032227,
|
| 1778 |
+
"loss": 0.1535,
|
| 1779 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1780 |
+
"rewards/chosen": -5.9134440422058105,
|
| 1781 |
+
"rewards/margins": 8.338058471679688,
|
| 1782 |
+
"rewards/rejected": -14.251502990722656,
|
| 1783 |
+
"step": 980
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 0.8508809626128062,
|
| 1787 |
+
"grad_norm": 2.881063461303711,
|
| 1788 |
+
"learning_rate": 1.2956158147457116e-06,
|
| 1789 |
+
"logits/chosen": 2.490622043609619,
|
| 1790 |
+
"logits/rejected": 2.438882350921631,
|
| 1791 |
+
"logps/chosen": -4.083024501800537,
|
| 1792 |
+
"logps/rejected": -8.853937149047852,
|
| 1793 |
+
"loss": 0.2337,
|
| 1794 |
+
"rewards/accuracies": 0.9375,
|
| 1795 |
+
"rewards/chosen": -6.124536991119385,
|
| 1796 |
+
"rewards/margins": 7.156369686126709,
|
| 1797 |
+
"rewards/rejected": -13.280906677246094,
|
| 1798 |
+
"step": 990
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 0.8594757198109153,
|
| 1802 |
+
"grad_norm": 2.4361555576324463,
|
| 1803 |
+
"learning_rate": 1.2500000000000007e-06,
|
| 1804 |
+
"logits/chosen": 2.5789554119110107,
|
| 1805 |
+
"logits/rejected": 2.755476951599121,
|
| 1806 |
+
"logps/chosen": -3.6121764183044434,
|
| 1807 |
+
"logps/rejected": -7.818983554840088,
|
| 1808 |
+
"loss": 0.2062,
|
| 1809 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1810 |
+
"rewards/chosen": -5.418264389038086,
|
| 1811 |
+
"rewards/margins": 6.310210227966309,
|
| 1812 |
+
"rewards/rejected": -11.728475570678711,
|
| 1813 |
+
"step": 1000
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 0.8594757198109153,
|
| 1817 |
+
"eval_logits/chosen": 2.3194375038146973,
|
| 1818 |
+
"eval_logits/rejected": 2.6192238330841064,
|
| 1819 |
+
"eval_logps/chosen": -3.992403745651245,
|
| 1820 |
+
"eval_logps/rejected": -9.769493103027344,
|
| 1821 |
+
"eval_loss": 0.2193986475467682,
|
| 1822 |
+
"eval_rewards/accuracies": 0.9473684430122375,
|
| 1823 |
+
"eval_rewards/chosen": -5.98860502243042,
|
| 1824 |
+
"eval_rewards/margins": 8.665634155273438,
|
| 1825 |
+
"eval_rewards/rejected": -14.654237747192383,
|
| 1826 |
+
"eval_runtime": 25.4033,
|
| 1827 |
+
"eval_samples_per_second": 29.642,
|
| 1828 |
+
"eval_steps_per_second": 3.74,
|
| 1829 |
+
"step": 1000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.8680704770090245,
|
| 1833 |
+
"grad_norm": 4.216193199157715,
|
| 1834 |
+
"learning_rate": 1.204932476567175e-06,
|
| 1835 |
+
"logits/chosen": 2.592757225036621,
|
| 1836 |
+
"logits/rejected": 2.777765989303589,
|
| 1837 |
+
"logps/chosen": -4.057430267333984,
|
| 1838 |
+
"logps/rejected": -8.371369361877441,
|
| 1839 |
+
"loss": 0.2322,
|
| 1840 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1841 |
+
"rewards/chosen": -6.086144924163818,
|
| 1842 |
+
"rewards/margins": 6.470907688140869,
|
| 1843 |
+
"rewards/rejected": -12.557052612304688,
|
| 1844 |
+
"step": 1010
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.8766652342071336,
|
| 1848 |
+
"grad_norm": 11.027617454528809,
|
| 1849 |
+
"learning_rate": 1.160433012552508e-06,
|
| 1850 |
+
"logits/chosen": 2.89387845993042,
|
| 1851 |
+
"logits/rejected": 2.8747739791870117,
|
| 1852 |
+
"logps/chosen": -4.125360012054443,
|
| 1853 |
+
"logps/rejected": -8.518974304199219,
|
| 1854 |
+
"loss": 0.2383,
|
| 1855 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1856 |
+
"rewards/chosen": -6.188040256500244,
|
| 1857 |
+
"rewards/margins": 6.590420722961426,
|
| 1858 |
+
"rewards/rejected": -12.778460502624512,
|
| 1859 |
+
"step": 1020
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.8852599914052428,
|
| 1863 |
+
"grad_norm": 3.2465028762817383,
|
| 1864 |
+
"learning_rate": 1.11652112689164e-06,
|
| 1865 |
+
"logits/chosen": 2.5261311531066895,
|
| 1866 |
+
"logits/rejected": 2.5254740715026855,
|
| 1867 |
+
"logps/chosen": -4.115787506103516,
|
| 1868 |
+
"logps/rejected": -9.08761215209961,
|
| 1869 |
+
"loss": 0.2262,
|
| 1870 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1871 |
+
"rewards/chosen": -6.173680782318115,
|
| 1872 |
+
"rewards/margins": 7.457736968994141,
|
| 1873 |
+
"rewards/rejected": -13.631416320800781,
|
| 1874 |
+
"step": 1030
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 0.8938547486033519,
|
| 1878 |
+
"grad_norm": 3.501631259918213,
|
| 1879 |
+
"learning_rate": 1.073216080788921e-06,
|
| 1880 |
+
"logits/chosen": 2.3466925621032715,
|
| 1881 |
+
"logits/rejected": 2.311033248901367,
|
| 1882 |
+
"logps/chosen": -3.7890372276306152,
|
| 1883 |
+
"logps/rejected": -8.364242553710938,
|
| 1884 |
+
"loss": 0.1909,
|
| 1885 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1886 |
+
"rewards/chosen": -5.683555603027344,
|
| 1887 |
+
"rewards/margins": 6.8628082275390625,
|
| 1888 |
+
"rewards/rejected": -12.546364784240723,
|
| 1889 |
+
"step": 1040
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 0.9024495058014611,
|
| 1893 |
+
"grad_norm": 2.803072690963745,
|
| 1894 |
+
"learning_rate": 1.0305368692688175e-06,
|
| 1895 |
+
"logits/chosen": 2.7137346267700195,
|
| 1896 |
+
"logits/rejected": 2.825568675994873,
|
| 1897 |
+
"logps/chosen": -3.8556430339813232,
|
| 1898 |
+
"logps/rejected": -8.831448554992676,
|
| 1899 |
+
"loss": 0.1965,
|
| 1900 |
+
"rewards/accuracies": 0.9375,
|
| 1901 |
+
"rewards/chosen": -5.783464431762695,
|
| 1902 |
+
"rewards/margins": 7.463706970214844,
|
| 1903 |
+
"rewards/rejected": -13.247172355651855,
|
| 1904 |
+
"step": 1050
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 0.9024495058014611,
|
| 1908 |
+
"eval_logits/chosen": 2.3855948448181152,
|
| 1909 |
+
"eval_logits/rejected": 2.6855878829956055,
|
| 1910 |
+
"eval_logps/chosen": -4.08680534362793,
|
| 1911 |
+
"eval_logps/rejected": -9.835321426391602,
|
| 1912 |
+
"eval_loss": 0.21824777126312256,
|
| 1913 |
+
"eval_rewards/accuracies": 0.9473684430122375,
|
| 1914 |
+
"eval_rewards/chosen": -6.130208492279053,
|
| 1915 |
+
"eval_rewards/margins": 8.622772216796875,
|
| 1916 |
+
"eval_rewards/rejected": -14.752982139587402,
|
| 1917 |
+
"eval_runtime": 25.407,
|
| 1918 |
+
"eval_samples_per_second": 29.637,
|
| 1919 |
+
"eval_steps_per_second": 3.739,
|
| 1920 |
+
"step": 1050
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.9110442629995703,
|
| 1924 |
+
"grad_norm": 3.8827121257781982,
|
| 1925 |
+
"learning_rate": 9.88502212844063e-07,
|
| 1926 |
+
"logits/chosen": 2.339254379272461,
|
| 1927 |
+
"logits/rejected": 2.321803092956543,
|
| 1928 |
+
"logps/chosen": -4.195307731628418,
|
| 1929 |
+
"logps/rejected": -8.984090805053711,
|
| 1930 |
+
"loss": 0.2003,
|
| 1931 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1932 |
+
"rewards/chosen": -6.292961120605469,
|
| 1933 |
+
"rewards/margins": 7.183175086975098,
|
| 1934 |
+
"rewards/rejected": -13.47613525390625,
|
| 1935 |
+
"step": 1060
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.9196390201976794,
|
| 1939 |
+
"grad_norm": 2.1895060539245605,
|
| 1940 |
+
"learning_rate": 9.471305493042243e-07,
|
| 1941 |
+
"logits/chosen": 2.5805039405822754,
|
| 1942 |
+
"logits/rejected": 2.725888729095459,
|
| 1943 |
+
"logps/chosen": -3.9197421073913574,
|
| 1944 |
+
"logps/rejected": -9.688833236694336,
|
| 1945 |
+
"loss": 0.1979,
|
| 1946 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1947 |
+
"rewards/chosen": -5.879612922668457,
|
| 1948 |
+
"rewards/margins": 8.653637886047363,
|
| 1949 |
+
"rewards/rejected": -14.533251762390137,
|
| 1950 |
+
"step": 1070
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 0.9282337773957886,
|
| 1954 |
+
"grad_norm": 2.7553319931030273,
|
| 1955 |
+
"learning_rate": 9.064400256282757e-07,
|
| 1956 |
+
"logits/chosen": 2.130366563796997,
|
| 1957 |
+
"logits/rejected": 2.3263564109802246,
|
| 1958 |
+
"logps/chosen": -4.264749526977539,
|
| 1959 |
+
"logps/rejected": -9.922590255737305,
|
| 1960 |
+
"loss": 0.1787,
|
| 1961 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1962 |
+
"rewards/chosen": -6.397124290466309,
|
| 1963 |
+
"rewards/margins": 8.486761093139648,
|
| 1964 |
+
"rewards/rejected": -14.883886337280273,
|
| 1965 |
+
"step": 1080
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 0.9368285345938977,
|
| 1969 |
+
"grad_norm": 3.2198596000671387,
|
| 1970 |
+
"learning_rate": 8.664484900247363e-07,
|
| 1971 |
+
"logits/chosen": 2.867159128189087,
|
| 1972 |
+
"logits/rejected": 2.5984394550323486,
|
| 1973 |
+
"logps/chosen": -3.8312149047851562,
|
| 1974 |
+
"logps/rejected": -8.639516830444336,
|
| 1975 |
+
"loss": 0.1971,
|
| 1976 |
+
"rewards/accuracies": 0.9375,
|
| 1977 |
+
"rewards/chosen": -5.746821403503418,
|
| 1978 |
+
"rewards/margins": 7.212453365325928,
|
| 1979 |
+
"rewards/rejected": -12.959274291992188,
|
| 1980 |
+
"step": 1090
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 0.9454232917920069,
|
| 1984 |
+
"grad_norm": 2.1565632820129395,
|
| 1985 |
+
"learning_rate": 8.271734841028553e-07,
|
| 1986 |
+
"logits/chosen": 3.2416539192199707,
|
| 1987 |
+
"logits/rejected": 2.98675274848938,
|
| 1988 |
+
"logps/chosen": -3.9062061309814453,
|
| 1989 |
+
"logps/rejected": -9.09926700592041,
|
| 1990 |
+
"loss": 0.1766,
|
| 1991 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1992 |
+
"rewards/chosen": -5.859309673309326,
|
| 1993 |
+
"rewards/margins": 7.789591312408447,
|
| 1994 |
+
"rewards/rejected": -13.648900032043457,
|
| 1995 |
+
"step": 1100
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.9454232917920069,
|
| 1999 |
+
"eval_logits/chosen": 2.452310562133789,
|
| 2000 |
+
"eval_logits/rejected": 2.728320837020874,
|
| 2001 |
+
"eval_logps/chosen": -4.018543243408203,
|
| 2002 |
+
"eval_logps/rejected": -9.840487480163574,
|
| 2003 |
+
"eval_loss": 0.21524910628795624,
|
| 2004 |
+
"eval_rewards/accuracies": 0.9473684430122375,
|
| 2005 |
+
"eval_rewards/chosen": -6.027814865112305,
|
| 2006 |
+
"eval_rewards/margins": 8.732914924621582,
|
| 2007 |
+
"eval_rewards/rejected": -14.760730743408203,
|
| 2008 |
+
"eval_runtime": 25.4021,
|
| 2009 |
+
"eval_samples_per_second": 29.643,
|
| 2010 |
+
"eval_steps_per_second": 3.74,
|
| 2011 |
+
"step": 1100
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.954018048990116,
|
| 2015 |
+
"grad_norm": 4.705589294433594,
|
| 2016 |
+
"learning_rate": 7.886322351782782e-07,
|
| 2017 |
+
"logits/chosen": 3.07834529876709,
|
| 2018 |
+
"logits/rejected": 2.766554117202759,
|
| 2019 |
+
"logps/chosen": -4.105443477630615,
|
| 2020 |
+
"logps/rejected": -8.913984298706055,
|
| 2021 |
+
"loss": 0.2274,
|
| 2022 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2023 |
+
"rewards/chosen": -6.158164978027344,
|
| 2024 |
+
"rewards/margins": 7.212811470031738,
|
| 2025 |
+
"rewards/rejected": -13.370976448059082,
|
| 2026 |
+
"step": 1110
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.9626128061882252,
|
| 2030 |
+
"grad_norm": 1.656952977180481,
|
| 2031 |
+
"learning_rate": 7.508416487165862e-07,
|
| 2032 |
+
"logits/chosen": 2.760190725326538,
|
| 2033 |
+
"logits/rejected": 2.4847655296325684,
|
| 2034 |
+
"logps/chosen": -3.7116692066192627,
|
| 2035 |
+
"logps/rejected": -8.578731536865234,
|
| 2036 |
+
"loss": 0.1609,
|
| 2037 |
+
"rewards/accuracies": 0.9375,
|
| 2038 |
+
"rewards/chosen": -5.567503452301025,
|
| 2039 |
+
"rewards/margins": 7.300595283508301,
|
| 2040 |
+
"rewards/rejected": -12.868098258972168,
|
| 2041 |
+
"step": 1120
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.9712075633863343,
|
| 2045 |
+
"grad_norm": 6.795106410980225,
|
| 2046 |
+
"learning_rate": 7.138183009179922e-07,
|
| 2047 |
+
"logits/chosen": 2.71061635017395,
|
| 2048 |
+
"logits/rejected": 2.543354034423828,
|
| 2049 |
+
"logps/chosen": -4.008772373199463,
|
| 2050 |
+
"logps/rejected": -9.546417236328125,
|
| 2051 |
+
"loss": 0.1813,
|
| 2052 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 2053 |
+
"rewards/chosen": -6.013156890869141,
|
| 2054 |
+
"rewards/margins": 8.306467056274414,
|
| 2055 |
+
"rewards/rejected": -14.319625854492188,
|
| 2056 |
+
"step": 1130
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 0.9798023205844435,
|
| 2060 |
+
"grad_norm": 3.892643690109253,
|
| 2061 |
+
"learning_rate": 6.775784314464717e-07,
|
| 2062 |
+
"logits/chosen": 2.826934814453125,
|
| 2063 |
+
"logits/rejected": 2.82672381401062,
|
| 2064 |
+
"logps/chosen": -3.863964796066284,
|
| 2065 |
+
"logps/rejected": -9.300487518310547,
|
| 2066 |
+
"loss": 0.1841,
|
| 2067 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2068 |
+
"rewards/chosen": -5.7959465980529785,
|
| 2069 |
+
"rewards/margins": 8.154784202575684,
|
| 2070 |
+
"rewards/rejected": -13.950732231140137,
|
| 2071 |
+
"step": 1140
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 0.9883970777825526,
|
| 2075 |
+
"grad_norm": 2.7565932273864746,
|
| 2076 |
+
"learning_rate": 6.421379363065142e-07,
|
| 2077 |
+
"logits/chosen": 2.8669843673706055,
|
| 2078 |
+
"logits/rejected": 2.703218936920166,
|
| 2079 |
+
"logps/chosen": -4.042016983032227,
|
| 2080 |
+
"logps/rejected": -9.321676254272461,
|
| 2081 |
+
"loss": 0.1698,
|
| 2082 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2083 |
+
"rewards/chosen": -6.06302547454834,
|
| 2084 |
+
"rewards/margins": 7.919488430023193,
|
| 2085 |
+
"rewards/rejected": -13.982513427734375,
|
| 2086 |
+
"step": 1150
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 0.9883970777825526,
|
| 2090 |
+
"eval_logits/chosen": 2.4691615104675293,
|
| 2091 |
+
"eval_logits/rejected": 2.7426767349243164,
|
| 2092 |
+
"eval_logps/chosen": -4.0566277503967285,
|
| 2093 |
+
"eval_logps/rejected": -9.914444923400879,
|
| 2094 |
+
"eval_loss": 0.2154170125722885,
|
| 2095 |
+
"eval_rewards/accuracies": 0.9473684430122375,
|
| 2096 |
+
"eval_rewards/chosen": -6.084941864013672,
|
| 2097 |
+
"eval_rewards/margins": 8.786724090576172,
|
| 2098 |
+
"eval_rewards/rejected": -14.871667861938477,
|
| 2099 |
+
"eval_runtime": 25.4309,
|
| 2100 |
+
"eval_samples_per_second": 29.61,
|
| 2101 |
+
"eval_steps_per_second": 3.736,
|
| 2102 |
+
"step": 1150
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.9969918349806618,
|
| 2106 |
+
"grad_norm": 5.313329219818115,
|
| 2107 |
+
"learning_rate": 6.075123608706093e-07,
|
| 2108 |
+
"logits/chosen": 2.488896131515503,
|
| 2109 |
+
"logits/rejected": 2.602234363555908,
|
| 2110 |
+
"logps/chosen": -4.304905891418457,
|
| 2111 |
+
"logps/rejected": -8.938867568969727,
|
| 2112 |
+
"loss": 0.187,
|
| 2113 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 2114 |
+
"rewards/chosen": -6.457359313964844,
|
| 2115 |
+
"rewards/margins": 6.9509429931640625,
|
| 2116 |
+
"rewards/rejected": -13.408302307128906,
|
| 2117 |
+
"step": 1160
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 1.0051568543188656,
|
| 2121 |
+
"grad_norm": 4.5972089767456055,
|
| 2122 |
+
"learning_rate": 5.737168930605272e-07,
|
| 2123 |
+
"logits/chosen": 2.951173782348633,
|
| 2124 |
+
"logits/rejected": 2.7277865409851074,
|
| 2125 |
+
"logps/chosen": -3.993422746658325,
|
| 2126 |
+
"logps/rejected": -9.191303253173828,
|
| 2127 |
+
"loss": 0.1904,
|
| 2128 |
+
"rewards/accuracies": 0.9342105388641357,
|
| 2129 |
+
"rewards/chosen": -5.990133285522461,
|
| 2130 |
+
"rewards/margins": 7.796821117401123,
|
| 2131 |
+
"rewards/rejected": -13.786954879760742,
|
| 2132 |
+
"step": 1170
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 1.0137516115169747,
|
| 2136 |
+
"grad_norm": 8.393081665039062,
|
| 2137 |
+
"learning_rate": 5.407663566854008e-07,
|
| 2138 |
+
"logits/chosen": 2.6279826164245605,
|
| 2139 |
+
"logits/rejected": 2.718173027038574,
|
| 2140 |
+
"logps/chosen": -4.164522647857666,
|
| 2141 |
+
"logps/rejected": -9.509660720825195,
|
| 2142 |
+
"loss": 0.1984,
|
| 2143 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2144 |
+
"rewards/chosen": -6.246783256530762,
|
| 2145 |
+
"rewards/margins": 8.017707824707031,
|
| 2146 |
+
"rewards/rejected": -14.264490127563477,
|
| 2147 |
+
"step": 1180
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 1.0223463687150838,
|
| 2151 |
+
"grad_norm": 1.666337251663208,
|
| 2152 |
+
"learning_rate": 5.086752049395094e-07,
|
| 2153 |
+
"logits/chosen": 3.449803590774536,
|
| 2154 |
+
"logits/rejected": 3.336327075958252,
|
| 2155 |
+
"logps/chosen": -4.214791297912598,
|
| 2156 |
+
"logps/rejected": -9.531986236572266,
|
| 2157 |
+
"loss": 0.1891,
|
| 2158 |
+
"rewards/accuracies": 0.925000011920929,
|
| 2159 |
+
"rewards/chosen": -6.322187423706055,
|
| 2160 |
+
"rewards/margins": 7.975789546966553,
|
| 2161 |
+
"rewards/rejected": -14.297978401184082,
|
| 2162 |
+
"step": 1190
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 1.0309411259131929,
|
| 2166 |
+
"grad_norm": 4.041133403778076,
|
| 2167 |
+
"learning_rate": 4.774575140626317e-07,
|
| 2168 |
+
"logits/chosen": 3.2861239910125732,
|
| 2169 |
+
"logits/rejected": 3.2246017456054688,
|
| 2170 |
+
"logps/chosen": -4.407445430755615,
|
| 2171 |
+
"logps/rejected": -9.031600952148438,
|
| 2172 |
+
"loss": 0.2037,
|
| 2173 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 2174 |
+
"rewards/chosen": -6.611169338226318,
|
| 2175 |
+
"rewards/margins": 6.936232566833496,
|
| 2176 |
+
"rewards/rejected": -13.547401428222656,
|
| 2177 |
+
"step": 1200
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 1.0309411259131929,
|
| 2181 |
+
"eval_logits/chosen": 2.4687464237213135,
|
| 2182 |
+
"eval_logits/rejected": 2.747849941253662,
|
| 2183 |
+
"eval_logps/chosen": -4.021546840667725,
|
| 2184 |
+
"eval_logps/rejected": -9.913942337036133,
|
| 2185 |
+
"eval_loss": 0.21217840909957886,
|
| 2186 |
+
"eval_rewards/accuracies": 0.9473684430122375,
|
| 2187 |
+
"eval_rewards/chosen": -6.032320499420166,
|
| 2188 |
+
"eval_rewards/margins": 8.838594436645508,
|
| 2189 |
+
"eval_rewards/rejected": -14.870915412902832,
|
| 2190 |
+
"eval_runtime": 25.4113,
|
| 2191 |
+
"eval_samples_per_second": 29.632,
|
| 2192 |
+
"eval_steps_per_second": 3.738,
|
| 2193 |
+
"step": 1200
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.0395358831113022,
|
| 2197 |
+
"grad_norm": 2.6725575923919678,
|
| 2198 |
+
"learning_rate": 4.4712697716573994e-07,
|
| 2199 |
+
"logits/chosen": 2.674346446990967,
|
| 2200 |
+
"logits/rejected": 2.491318941116333,
|
| 2201 |
+
"logps/chosen": -3.7565529346466064,
|
| 2202 |
+
"logps/rejected": -7.836224555969238,
|
| 2203 |
+
"loss": 0.2035,
|
| 2204 |
+
"rewards/accuracies": 0.9375,
|
| 2205 |
+
"rewards/chosen": -5.634829521179199,
|
| 2206 |
+
"rewards/margins": 6.119507789611816,
|
| 2207 |
+
"rewards/rejected": -11.7543363571167,
|
| 2208 |
+
"step": 1210
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 1.0481306403094113,
|
| 2212 |
+
"grad_norm": 3.706759214401245,
|
| 2213 |
+
"learning_rate": 4.1769689822475147e-07,
|
| 2214 |
+
"logits/chosen": 2.4985451698303223,
|
| 2215 |
+
"logits/rejected": 2.393381118774414,
|
| 2216 |
+
"logps/chosen": -3.8855807781219482,
|
| 2217 |
+
"logps/rejected": -8.915465354919434,
|
| 2218 |
+
"loss": 0.192,
|
| 2219 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2220 |
+
"rewards/chosen": -5.828372001647949,
|
| 2221 |
+
"rewards/margins": 7.544825553894043,
|
| 2222 |
+
"rewards/rejected": -13.373197555541992,
|
| 2223 |
+
"step": 1220
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 1.0567253975075204,
|
| 2227 |
+
"grad_norm": 3.47476863861084,
|
| 2228 |
+
"learning_rate": 3.891801862449629e-07,
|
| 2229 |
+
"logits/chosen": 2.4244041442871094,
|
| 2230 |
+
"logits/rejected": 2.176212787628174,
|
| 2231 |
+
"logps/chosen": -4.165676593780518,
|
| 2232 |
+
"logps/rejected": -9.606474876403809,
|
| 2233 |
+
"loss": 0.1734,
|
| 2234 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2235 |
+
"rewards/chosen": -6.2485151290893555,
|
| 2236 |
+
"rewards/margins": 8.1611967086792,
|
| 2237 |
+
"rewards/rejected": -14.409710884094238,
|
| 2238 |
+
"step": 1230
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 1.0653201547056295,
|
| 2242 |
+
"grad_norm": 7.537776470184326,
|
| 2243 |
+
"learning_rate": 3.615893495987335e-07,
|
| 2244 |
+
"logits/chosen": 3.5634818077087402,
|
| 2245 |
+
"logits/rejected": 3.599396228790283,
|
| 2246 |
+
"logps/chosen": -3.8801186084747314,
|
| 2247 |
+
"logps/rejected": -8.772649765014648,
|
| 2248 |
+
"loss": 0.1907,
|
| 2249 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2250 |
+
"rewards/chosen": -5.8201775550842285,
|
| 2251 |
+
"rewards/margins": 7.338797092437744,
|
| 2252 |
+
"rewards/rejected": -13.158975601196289,
|
| 2253 |
+
"step": 1240
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 1.0739149119037388,
|
| 2257 |
+
"grad_norm": 7.371191501617432,
|
| 2258 |
+
"learning_rate": 3.3493649053890325e-07,
|
| 2259 |
+
"logits/chosen": 1.536439299583435,
|
| 2260 |
+
"logits/rejected": 1.6537296772003174,
|
| 2261 |
+
"logps/chosen": -3.4690723419189453,
|
| 2262 |
+
"logps/rejected": -9.245938301086426,
|
| 2263 |
+
"loss": 0.1963,
|
| 2264 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2265 |
+
"rewards/chosen": -5.203608512878418,
|
| 2266 |
+
"rewards/margins": 8.665300369262695,
|
| 2267 |
+
"rewards/rejected": -13.868908882141113,
|
| 2268 |
+
"step": 1250
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 1.0739149119037388,
|
| 2272 |
+
"eval_logits/chosen": 2.463451862335205,
|
| 2273 |
+
"eval_logits/rejected": 2.7573211193084717,
|
| 2274 |
+
"eval_logps/chosen": -4.089573383331299,
|
| 2275 |
+
"eval_logps/rejected": -10.03561019897461,
|
| 2276 |
+
"eval_loss": 0.21280600130558014,
|
| 2277 |
+
"eval_rewards/accuracies": 0.9473684430122375,
|
| 2278 |
+
"eval_rewards/chosen": -6.1343607902526855,
|
| 2279 |
+
"eval_rewards/margins": 8.919055938720703,
|
| 2280 |
+
"eval_rewards/rejected": -15.05341625213623,
|
| 2281 |
+
"eval_runtime": 26.0304,
|
| 2282 |
+
"eval_samples_per_second": 28.928,
|
| 2283 |
+
"eval_steps_per_second": 3.65,
|
| 2284 |
+
"step": 1250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 1.0825096691018479,
|
| 2288 |
+
"grad_norm": 3.3083207607269287,
|
| 2289 |
+
"learning_rate": 3.092332998903416e-07,
|
| 2290 |
+
"logits/chosen": 2.6976983547210693,
|
| 2291 |
+
"logits/rejected": 2.558777332305908,
|
| 2292 |
+
"logps/chosen": -4.322577953338623,
|
| 2293 |
+
"logps/rejected": -9.3988037109375,
|
| 2294 |
+
"loss": 0.1782,
|
| 2295 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2296 |
+
"rewards/chosen": -6.483867645263672,
|
| 2297 |
+
"rewards/margins": 7.614336967468262,
|
| 2298 |
+
"rewards/rejected": -14.098203659057617,
|
| 2299 |
+
"step": 1260
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 1.091104426299957,
|
| 2303 |
+
"grad_norm": 2.7674386501312256,
|
| 2304 |
+
"learning_rate": 2.844910519219632e-07,
|
| 2305 |
+
"logits/chosen": 2.5444347858428955,
|
| 2306 |
+
"logits/rejected": 2.5096020698547363,
|
| 2307 |
+
"logps/chosen": -4.426403522491455,
|
| 2308 |
+
"logps/rejected": -9.394776344299316,
|
| 2309 |
+
"loss": 0.1731,
|
| 2310 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 2311 |
+
"rewards/chosen": -6.639605522155762,
|
| 2312 |
+
"rewards/margins": 7.4525580406188965,
|
| 2313 |
+
"rewards/rejected": -14.0921630859375,
|
| 2314 |
+
"step": 1270
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 1.0996991834980663,
|
| 2318 |
+
"grad_norm": 3.5264317989349365,
|
| 2319 |
+
"learning_rate": 2.6072059940146775e-07,
|
| 2320 |
+
"logits/chosen": 2.0977225303649902,
|
| 2321 |
+
"logits/rejected": 2.335002899169922,
|
| 2322 |
+
"logps/chosen": -4.064654350280762,
|
| 2323 |
+
"logps/rejected": -9.52751350402832,
|
| 2324 |
+
"loss": 0.1904,
|
| 2325 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2326 |
+
"rewards/chosen": -6.096982002258301,
|
| 2327 |
+
"rewards/margins": 8.19428825378418,
|
| 2328 |
+
"rewards/rejected": -14.291269302368164,
|
| 2329 |
+
"step": 1280
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 1.1082939406961754,
|
| 2333 |
+
"grad_norm": 3.810760498046875,
|
| 2334 |
+
"learning_rate": 2.3793236883495164e-07,
|
| 2335 |
+
"logits/chosen": 2.6432700157165527,
|
| 2336 |
+
"logits/rejected": 2.542976140975952,
|
| 2337 |
+
"logps/chosen": -4.272869110107422,
|
| 2338 |
+
"logps/rejected": -9.520709037780762,
|
| 2339 |
+
"loss": 0.1787,
|
| 2340 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2341 |
+
"rewards/chosen": -6.409303188323975,
|
| 2342 |
+
"rewards/margins": 7.871760368347168,
|
| 2343 |
+
"rewards/rejected": -14.281064987182617,
|
| 2344 |
+
"step": 1290
|
| 2345 |
+
},
|
| 2346 |
+
{
|
| 2347 |
+
"epoch": 1.1168886978942845,
|
| 2348 |
+
"grad_norm": 5.784368991851807,
|
| 2349 |
+
"learning_rate": 2.1613635589349756e-07,
|
| 2350 |
+
"logits/chosen": 2.5322649478912354,
|
| 2351 |
+
"logits/rejected": 2.5456764698028564,
|
| 2352 |
+
"logps/chosen": -4.310420513153076,
|
| 2353 |
+
"logps/rejected": -8.846685409545898,
|
| 2354 |
+
"loss": 0.2135,
|
| 2355 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2356 |
+
"rewards/chosen": -6.465630531311035,
|
| 2357 |
+
"rewards/margins": 6.8043975830078125,
|
| 2358 |
+
"rewards/rejected": -13.270029067993164,
|
| 2359 |
+
"step": 1300
|
| 2360 |
+
},
|
| 2361 |
+
{
|
| 2362 |
+
"epoch": 1.1168886978942845,
|
| 2363 |
+
"eval_logits/chosen": 2.43306040763855,
|
| 2364 |
+
"eval_logits/rejected": 2.7410776615142822,
|
| 2365 |
+
"eval_logps/chosen": -4.157689571380615,
|
| 2366 |
+
"eval_logps/rejected": -10.128125190734863,
|
| 2367 |
+
"eval_loss": 0.21243833005428314,
|
| 2368 |
+
"eval_rewards/accuracies": 0.9473684430122375,
|
| 2369 |
+
"eval_rewards/chosen": -6.236534595489502,
|
| 2370 |
+
"eval_rewards/margins": 8.95565414428711,
|
| 2371 |
+
"eval_rewards/rejected": -15.192190170288086,
|
| 2372 |
+
"eval_runtime": 25.4225,
|
| 2373 |
+
"eval_samples_per_second": 29.619,
|
| 2374 |
+
"eval_steps_per_second": 3.737,
|
| 2375 |
+
"step": 1300
|
| 2376 |
+
}
|
| 2377 |
+
],
|
| 2378 |
+
"logging_steps": 10,
|
| 2379 |
+
"max_steps": 1500,
|
| 2380 |
+
"num_input_tokens_seen": 0,
|
| 2381 |
+
"num_train_epochs": 2,
|
| 2382 |
+
"save_steps": 50,
|
| 2383 |
+
"stateful_callbacks": {
|
| 2384 |
+
"TrainerControl": {
|
| 2385 |
+
"args": {
|
| 2386 |
+
"should_epoch_stop": false,
|
| 2387 |
+
"should_evaluate": false,
|
| 2388 |
+
"should_log": false,
|
| 2389 |
+
"should_save": true,
|
| 2390 |
+
"should_training_stop": false
|
| 2391 |
+
},
|
| 2392 |
+
"attributes": {}
|
| 2393 |
+
}
|
| 2394 |
+
},
|
| 2395 |
+
"total_flos": 2.9892761024819364e+18,
|
| 2396 |
+
"train_batch_size": 1,
|
| 2397 |
+
"trial_name": null,
|
| 2398 |
+
"trial_params": null
|
| 2399 |
+
}
|
checkpoint-1300/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0e56bd7e0b9c14ab415ac512838b27c755a593199eea9fac203b4183c2cbca02
|
| 3 |
+
size 7224
|
checkpoint-1300/zero_to_fp32.py
ADDED
|
@@ -0,0 +1,674 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
#!/usr/bin/env python
|
| 2 |
+
|
| 3 |
+
# Copyright (c) Microsoft Corporation.
|
| 4 |
+
# SPDX-License-Identifier: Apache-2.0
|
| 5 |
+
|
| 6 |
+
# DeepSpeed Team
|
| 7 |
+
|
| 8 |
+
# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
|
| 9 |
+
# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
|
| 10 |
+
# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
|
| 11 |
+
# application.
|
| 12 |
+
#
|
| 13 |
+
# example:
|
| 14 |
+
# python zero_to_fp32.py . output_dir/
|
| 15 |
+
# or
|
| 16 |
+
# python zero_to_fp32.py . output_dir/ --safe_serialization
|
| 17 |
+
|
| 18 |
+
import argparse
|
| 19 |
+
import torch
|
| 20 |
+
import glob
|
| 21 |
+
import math
|
| 22 |
+
import os
|
| 23 |
+
import re
|
| 24 |
+
import json
|
| 25 |
+
from tqdm import tqdm
|
| 26 |
+
from collections import OrderedDict
|
| 27 |
+
from dataclasses import dataclass
|
| 28 |
+
|
| 29 |
+
# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
|
| 30 |
+
# DeepSpeed data structures it has to be available in the current python environment.
|
| 31 |
+
from deepspeed.utils import logger
|
| 32 |
+
from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
|
| 33 |
+
FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
|
| 34 |
+
FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
|
| 35 |
+
|
| 36 |
+
|
| 37 |
+
@dataclass
|
| 38 |
+
class zero_model_state:
|
| 39 |
+
buffers: dict()
|
| 40 |
+
param_shapes: dict()
|
| 41 |
+
shared_params: list
|
| 42 |
+
ds_version: int
|
| 43 |
+
frozen_param_shapes: dict()
|
| 44 |
+
frozen_param_fragments: dict()
|
| 45 |
+
|
| 46 |
+
|
| 47 |
+
debug = 0
|
| 48 |
+
|
| 49 |
+
# load to cpu
|
| 50 |
+
device = torch.device('cpu')
|
| 51 |
+
|
| 52 |
+
|
| 53 |
+
def atoi(text):
|
| 54 |
+
return int(text) if text.isdigit() else text
|
| 55 |
+
|
| 56 |
+
|
| 57 |
+
def natural_keys(text):
|
| 58 |
+
'''
|
| 59 |
+
alist.sort(key=natural_keys) sorts in human order
|
| 60 |
+
http://nedbatchelder.com/blog/200712/human_sorting.html
|
| 61 |
+
(See Toothy's implementation in the comments)
|
| 62 |
+
'''
|
| 63 |
+
return [atoi(c) for c in re.split(r'(\d+)', text)]
|
| 64 |
+
|
| 65 |
+
|
| 66 |
+
def get_model_state_file(checkpoint_dir, zero_stage):
|
| 67 |
+
if not os.path.isdir(checkpoint_dir):
|
| 68 |
+
raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
|
| 69 |
+
|
| 70 |
+
# there should be only one file
|
| 71 |
+
if zero_stage <= 2:
|
| 72 |
+
file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
|
| 73 |
+
elif zero_stage == 3:
|
| 74 |
+
file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
|
| 75 |
+
|
| 76 |
+
if not os.path.exists(file):
|
| 77 |
+
raise FileNotFoundError(f"can't find model states file at '{file}'")
|
| 78 |
+
|
| 79 |
+
return file
|
| 80 |
+
|
| 81 |
+
|
| 82 |
+
def get_checkpoint_files(checkpoint_dir, glob_pattern):
|
| 83 |
+
# XXX: need to test that this simple glob rule works for multi-node setup too
|
| 84 |
+
ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
|
| 85 |
+
|
| 86 |
+
if len(ckpt_files) == 0:
|
| 87 |
+
raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
|
| 88 |
+
|
| 89 |
+
return ckpt_files
|
| 90 |
+
|
| 91 |
+
|
| 92 |
+
def get_optim_files(checkpoint_dir):
|
| 93 |
+
return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
|
| 94 |
+
|
| 95 |
+
|
| 96 |
+
def get_model_state_files(checkpoint_dir):
|
| 97 |
+
return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
|
| 98 |
+
|
| 99 |
+
|
| 100 |
+
def parse_model_states(files):
|
| 101 |
+
zero_model_states = []
|
| 102 |
+
for file in files:
|
| 103 |
+
state_dict = torch.load(file, map_location=device)
|
| 104 |
+
|
| 105 |
+
if BUFFER_NAMES not in state_dict:
|
| 106 |
+
raise ValueError(f"{file} is not a model state checkpoint")
|
| 107 |
+
buffer_names = state_dict[BUFFER_NAMES]
|
| 108 |
+
if debug:
|
| 109 |
+
print("Found buffers:", buffer_names)
|
| 110 |
+
|
| 111 |
+
# recover just the buffers while restoring them to fp32 if they were saved in fp16
|
| 112 |
+
buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
|
| 113 |
+
param_shapes = state_dict[PARAM_SHAPES]
|
| 114 |
+
|
| 115 |
+
# collect parameters that are included in param_shapes
|
| 116 |
+
param_names = []
|
| 117 |
+
for s in param_shapes:
|
| 118 |
+
for name in s.keys():
|
| 119 |
+
param_names.append(name)
|
| 120 |
+
|
| 121 |
+
# update with frozen parameters
|
| 122 |
+
frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
|
| 123 |
+
if frozen_param_shapes is not None:
|
| 124 |
+
if debug:
|
| 125 |
+
print(f"Found frozen_param_shapes: {frozen_param_shapes}")
|
| 126 |
+
param_names += list(frozen_param_shapes.keys())
|
| 127 |
+
|
| 128 |
+
# handle shared params
|
| 129 |
+
shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
|
| 130 |
+
|
| 131 |
+
ds_version = state_dict.get(DS_VERSION, None)
|
| 132 |
+
|
| 133 |
+
frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
|
| 134 |
+
|
| 135 |
+
z_model_state = zero_model_state(buffers=buffers,
|
| 136 |
+
param_shapes=param_shapes,
|
| 137 |
+
shared_params=shared_params,
|
| 138 |
+
ds_version=ds_version,
|
| 139 |
+
frozen_param_shapes=frozen_param_shapes,
|
| 140 |
+
frozen_param_fragments=frozen_param_fragments)
|
| 141 |
+
zero_model_states.append(z_model_state)
|
| 142 |
+
|
| 143 |
+
return zero_model_states
|
| 144 |
+
|
| 145 |
+
|
| 146 |
+
def parse_optim_states(files, ds_checkpoint_dir):
|
| 147 |
+
total_files = len(files)
|
| 148 |
+
state_dicts = []
|
| 149 |
+
for f in files:
|
| 150 |
+
state_dict = torch.load(f, map_location=device)
|
| 151 |
+
# immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
|
| 152 |
+
# and also handle the case where it was already removed by another helper script
|
| 153 |
+
state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
|
| 154 |
+
state_dicts.append(state_dict)
|
| 155 |
+
|
| 156 |
+
if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
|
| 157 |
+
raise ValueError(f"{files[0]} is not a zero checkpoint")
|
| 158 |
+
zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
|
| 159 |
+
world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
|
| 160 |
+
|
| 161 |
+
# For ZeRO-2 each param group can have different partition_count as data parallelism for expert
|
| 162 |
+
# parameters can be different from data parallelism for non-expert parameters. So we can just
|
| 163 |
+
# use the max of the partition_count to get the dp world_size.
|
| 164 |
+
|
| 165 |
+
if type(world_size) is list:
|
| 166 |
+
world_size = max(world_size)
|
| 167 |
+
|
| 168 |
+
if world_size != total_files:
|
| 169 |
+
raise ValueError(
|
| 170 |
+
f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
|
| 171 |
+
"Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
|
| 172 |
+
)
|
| 173 |
+
|
| 174 |
+
# the groups are named differently in each stage
|
| 175 |
+
if zero_stage <= 2:
|
| 176 |
+
fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
|
| 177 |
+
elif zero_stage == 3:
|
| 178 |
+
fp32_groups_key = FP32_FLAT_GROUPS
|
| 179 |
+
else:
|
| 180 |
+
raise ValueError(f"unknown zero stage {zero_stage}")
|
| 181 |
+
|
| 182 |
+
if zero_stage <= 2:
|
| 183 |
+
fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
|
| 184 |
+
elif zero_stage == 3:
|
| 185 |
+
# if there is more than one param group, there will be multiple flattened tensors - one
|
| 186 |
+
# flattened tensor per group - for simplicity merge them into a single tensor
|
| 187 |
+
#
|
| 188 |
+
# XXX: could make the script more memory efficient for when there are multiple groups - it
|
| 189 |
+
# will require matching the sub-lists of param_shapes for each param group flattened tensor
|
| 190 |
+
|
| 191 |
+
fp32_flat_groups = [
|
| 192 |
+
torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
|
| 193 |
+
]
|
| 194 |
+
|
| 195 |
+
return zero_stage, world_size, fp32_flat_groups
|
| 196 |
+
|
| 197 |
+
|
| 198 |
+
def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
|
| 199 |
+
"""
|
| 200 |
+
Returns fp32 state_dict reconstructed from ds checkpoint
|
| 201 |
+
|
| 202 |
+
Args:
|
| 203 |
+
- ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
|
| 204 |
+
|
| 205 |
+
"""
|
| 206 |
+
print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
|
| 207 |
+
|
| 208 |
+
optim_files = get_optim_files(ds_checkpoint_dir)
|
| 209 |
+
zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
|
| 210 |
+
print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
|
| 211 |
+
|
| 212 |
+
model_files = get_model_state_files(ds_checkpoint_dir)
|
| 213 |
+
|
| 214 |
+
zero_model_states = parse_model_states(model_files)
|
| 215 |
+
print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
|
| 216 |
+
|
| 217 |
+
if zero_stage <= 2:
|
| 218 |
+
return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 219 |
+
exclude_frozen_parameters)
|
| 220 |
+
elif zero_stage == 3:
|
| 221 |
+
return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 222 |
+
exclude_frozen_parameters)
|
| 223 |
+
|
| 224 |
+
|
| 225 |
+
def _zero2_merge_frozen_params(state_dict, zero_model_states):
|
| 226 |
+
if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
|
| 227 |
+
return
|
| 228 |
+
|
| 229 |
+
frozen_param_shapes = zero_model_states[0].frozen_param_shapes
|
| 230 |
+
frozen_param_fragments = zero_model_states[0].frozen_param_fragments
|
| 231 |
+
|
| 232 |
+
if debug:
|
| 233 |
+
num_elem = sum(s.numel() for s in frozen_param_shapes.values())
|
| 234 |
+
print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
|
| 235 |
+
|
| 236 |
+
wanted_params = len(frozen_param_shapes)
|
| 237 |
+
wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
|
| 238 |
+
avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
|
| 239 |
+
print(f'Frozen params: Have {avail_numel} numels to process.')
|
| 240 |
+
print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
|
| 241 |
+
|
| 242 |
+
total_params = 0
|
| 243 |
+
total_numel = 0
|
| 244 |
+
for name, shape in frozen_param_shapes.items():
|
| 245 |
+
total_params += 1
|
| 246 |
+
unpartitioned_numel = shape.numel()
|
| 247 |
+
total_numel += unpartitioned_numel
|
| 248 |
+
|
| 249 |
+
state_dict[name] = frozen_param_fragments[name]
|
| 250 |
+
|
| 251 |
+
if debug:
|
| 252 |
+
print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
|
| 253 |
+
|
| 254 |
+
print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
|
| 255 |
+
|
| 256 |
+
|
| 257 |
+
def _has_callable(obj, fn):
|
| 258 |
+
attr = getattr(obj, fn, None)
|
| 259 |
+
return callable(attr)
|
| 260 |
+
|
| 261 |
+
|
| 262 |
+
def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
|
| 263 |
+
param_shapes = zero_model_states[0].param_shapes
|
| 264 |
+
|
| 265 |
+
# Reconstruction protocol:
|
| 266 |
+
#
|
| 267 |
+
# XXX: document this
|
| 268 |
+
|
| 269 |
+
if debug:
|
| 270 |
+
for i in range(world_size):
|
| 271 |
+
for j in range(len(fp32_flat_groups[0])):
|
| 272 |
+
print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
|
| 273 |
+
|
| 274 |
+
# XXX: memory usage doubles here (zero2)
|
| 275 |
+
num_param_groups = len(fp32_flat_groups[0])
|
| 276 |
+
merged_single_partition_of_fp32_groups = []
|
| 277 |
+
for i in range(num_param_groups):
|
| 278 |
+
merged_partitions = [sd[i] for sd in fp32_flat_groups]
|
| 279 |
+
full_single_fp32_vector = torch.cat(merged_partitions, 0)
|
| 280 |
+
merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
|
| 281 |
+
avail_numel = sum(
|
| 282 |
+
[full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
|
| 283 |
+
|
| 284 |
+
if debug:
|
| 285 |
+
wanted_params = sum([len(shapes) for shapes in param_shapes])
|
| 286 |
+
wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
|
| 287 |
+
# not asserting if there is a mismatch due to possible padding
|
| 288 |
+
print(f"Have {avail_numel} numels to process.")
|
| 289 |
+
print(f"Need {wanted_numel} numels in {wanted_params} params.")
|
| 290 |
+
|
| 291 |
+
# params
|
| 292 |
+
# XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
|
| 293 |
+
# out-of-core computing solution
|
| 294 |
+
total_numel = 0
|
| 295 |
+
total_params = 0
|
| 296 |
+
for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
|
| 297 |
+
offset = 0
|
| 298 |
+
avail_numel = full_single_fp32_vector.numel()
|
| 299 |
+
for name, shape in shapes.items():
|
| 300 |
+
|
| 301 |
+
unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
|
| 302 |
+
total_numel += unpartitioned_numel
|
| 303 |
+
total_params += 1
|
| 304 |
+
|
| 305 |
+
if debug:
|
| 306 |
+
print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
|
| 307 |
+
state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
|
| 308 |
+
offset += unpartitioned_numel
|
| 309 |
+
|
| 310 |
+
# Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
|
| 311 |
+
# avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
|
| 312 |
+
# paddings performed in the code it's almost impossible to predict the exact numbers w/o the
|
| 313 |
+
# live optimizer object, so we are checking that the numbers are within the right range
|
| 314 |
+
align_to = 2 * world_size
|
| 315 |
+
|
| 316 |
+
def zero2_align(x):
|
| 317 |
+
return align_to * math.ceil(x / align_to)
|
| 318 |
+
|
| 319 |
+
if debug:
|
| 320 |
+
print(f"original offset={offset}, avail_numel={avail_numel}")
|
| 321 |
+
|
| 322 |
+
offset = zero2_align(offset)
|
| 323 |
+
avail_numel = zero2_align(avail_numel)
|
| 324 |
+
|
| 325 |
+
if debug:
|
| 326 |
+
print(f"aligned offset={offset}, avail_numel={avail_numel}")
|
| 327 |
+
|
| 328 |
+
# Sanity check
|
| 329 |
+
if offset != avail_numel:
|
| 330 |
+
raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
|
| 331 |
+
|
| 332 |
+
print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
|
| 333 |
+
|
| 334 |
+
|
| 335 |
+
def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 336 |
+
exclude_frozen_parameters):
|
| 337 |
+
state_dict = OrderedDict()
|
| 338 |
+
|
| 339 |
+
# buffers
|
| 340 |
+
buffers = zero_model_states[0].buffers
|
| 341 |
+
state_dict.update(buffers)
|
| 342 |
+
if debug:
|
| 343 |
+
print(f"added {len(buffers)} buffers")
|
| 344 |
+
|
| 345 |
+
if not exclude_frozen_parameters:
|
| 346 |
+
_zero2_merge_frozen_params(state_dict, zero_model_states)
|
| 347 |
+
|
| 348 |
+
_zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
|
| 349 |
+
|
| 350 |
+
# recover shared parameters
|
| 351 |
+
for pair in zero_model_states[0].shared_params:
|
| 352 |
+
if pair[1] in state_dict:
|
| 353 |
+
state_dict[pair[0]] = state_dict[pair[1]]
|
| 354 |
+
|
| 355 |
+
return state_dict
|
| 356 |
+
|
| 357 |
+
|
| 358 |
+
def zero3_partitioned_param_info(unpartitioned_numel, world_size):
|
| 359 |
+
remainder = unpartitioned_numel % world_size
|
| 360 |
+
padding_numel = (world_size - remainder) if remainder else 0
|
| 361 |
+
partitioned_numel = math.ceil(unpartitioned_numel / world_size)
|
| 362 |
+
return partitioned_numel, padding_numel
|
| 363 |
+
|
| 364 |
+
|
| 365 |
+
def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
|
| 366 |
+
if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
|
| 367 |
+
return
|
| 368 |
+
|
| 369 |
+
if debug:
|
| 370 |
+
for i in range(world_size):
|
| 371 |
+
num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
|
| 372 |
+
print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
|
| 373 |
+
|
| 374 |
+
frozen_param_shapes = zero_model_states[0].frozen_param_shapes
|
| 375 |
+
wanted_params = len(frozen_param_shapes)
|
| 376 |
+
wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
|
| 377 |
+
avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
|
| 378 |
+
print(f'Frozen params: Have {avail_numel} numels to process.')
|
| 379 |
+
print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
|
| 380 |
+
|
| 381 |
+
total_params = 0
|
| 382 |
+
total_numel = 0
|
| 383 |
+
for name, shape in zero_model_states[0].frozen_param_shapes.items():
|
| 384 |
+
total_params += 1
|
| 385 |
+
unpartitioned_numel = shape.numel()
|
| 386 |
+
total_numel += unpartitioned_numel
|
| 387 |
+
|
| 388 |
+
param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
|
| 389 |
+
state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
|
| 390 |
+
|
| 391 |
+
partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
|
| 392 |
+
|
| 393 |
+
if debug:
|
| 394 |
+
print(
|
| 395 |
+
f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
|
| 396 |
+
)
|
| 397 |
+
|
| 398 |
+
print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
|
| 399 |
+
|
| 400 |
+
|
| 401 |
+
def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
|
| 402 |
+
param_shapes = zero_model_states[0].param_shapes
|
| 403 |
+
avail_numel = fp32_flat_groups[0].numel() * world_size
|
| 404 |
+
# Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
|
| 405 |
+
# param, re-consolidating each param, while dealing with padding if any
|
| 406 |
+
|
| 407 |
+
# merge list of dicts, preserving order
|
| 408 |
+
param_shapes = {k: v for d in param_shapes for k, v in d.items()}
|
| 409 |
+
|
| 410 |
+
if debug:
|
| 411 |
+
for i in range(world_size):
|
| 412 |
+
print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
|
| 413 |
+
|
| 414 |
+
wanted_params = len(param_shapes)
|
| 415 |
+
wanted_numel = sum(shape.numel() for shape in param_shapes.values())
|
| 416 |
+
# not asserting if there is a mismatch due to possible padding
|
| 417 |
+
avail_numel = fp32_flat_groups[0].numel() * world_size
|
| 418 |
+
print(f"Trainable params: Have {avail_numel} numels to process.")
|
| 419 |
+
print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
|
| 420 |
+
|
| 421 |
+
# params
|
| 422 |
+
# XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
|
| 423 |
+
# out-of-core computing solution
|
| 424 |
+
offset = 0
|
| 425 |
+
total_numel = 0
|
| 426 |
+
total_params = 0
|
| 427 |
+
for name, shape in tqdm(param_shapes.items(), desc='Gathering Sharded Weights'):
|
| 428 |
+
unpartitioned_numel = shape.numel()
|
| 429 |
+
total_numel += unpartitioned_numel
|
| 430 |
+
total_params += 1
|
| 431 |
+
partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
|
| 432 |
+
|
| 433 |
+
if debug:
|
| 434 |
+
print(
|
| 435 |
+
f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
|
| 436 |
+
)
|
| 437 |
+
|
| 438 |
+
# XXX: memory usage doubles here
|
| 439 |
+
state_dict[name] = torch.cat(
|
| 440 |
+
tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
|
| 441 |
+
0).narrow(0, 0, unpartitioned_numel).view(shape)
|
| 442 |
+
offset += partitioned_numel
|
| 443 |
+
|
| 444 |
+
offset *= world_size
|
| 445 |
+
|
| 446 |
+
# Sanity check
|
| 447 |
+
if offset != avail_numel:
|
| 448 |
+
raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
|
| 449 |
+
|
| 450 |
+
print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
|
| 451 |
+
|
| 452 |
+
|
| 453 |
+
def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 454 |
+
exclude_frozen_parameters):
|
| 455 |
+
state_dict = OrderedDict()
|
| 456 |
+
|
| 457 |
+
# buffers
|
| 458 |
+
buffers = zero_model_states[0].buffers
|
| 459 |
+
state_dict.update(buffers)
|
| 460 |
+
if debug:
|
| 461 |
+
print(f"added {len(buffers)} buffers")
|
| 462 |
+
|
| 463 |
+
if not exclude_frozen_parameters:
|
| 464 |
+
_zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
|
| 465 |
+
|
| 466 |
+
_zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
|
| 467 |
+
|
| 468 |
+
# recover shared parameters
|
| 469 |
+
for pair in zero_model_states[0].shared_params:
|
| 470 |
+
if pair[1] in state_dict:
|
| 471 |
+
state_dict[pair[0]] = state_dict[pair[1]]
|
| 472 |
+
|
| 473 |
+
return state_dict
|
| 474 |
+
|
| 475 |
+
|
| 476 |
+
def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False):
|
| 477 |
+
"""
|
| 478 |
+
Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
|
| 479 |
+
``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
|
| 480 |
+
via a model hub.
|
| 481 |
+
|
| 482 |
+
Args:
|
| 483 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder
|
| 484 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
|
| 485 |
+
- ``exclude_frozen_parameters``: exclude frozen parameters
|
| 486 |
+
|
| 487 |
+
Returns:
|
| 488 |
+
- pytorch ``state_dict``
|
| 489 |
+
|
| 490 |
+
Note: this approach may not work if your application doesn't have sufficient free CPU memory and
|
| 491 |
+
you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
|
| 492 |
+
the checkpoint.
|
| 493 |
+
|
| 494 |
+
A typical usage might be ::
|
| 495 |
+
|
| 496 |
+
from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
|
| 497 |
+
# do the training and checkpoint saving
|
| 498 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
|
| 499 |
+
model = model.cpu() # move to cpu
|
| 500 |
+
model.load_state_dict(state_dict)
|
| 501 |
+
# submit to model hub or save the model to share with others
|
| 502 |
+
|
| 503 |
+
In this example the ``model`` will no longer be usable in the deepspeed context of the same
|
| 504 |
+
application. i.e. you will need to re-initialize the deepspeed engine, since
|
| 505 |
+
``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
|
| 506 |
+
|
| 507 |
+
If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
|
| 508 |
+
|
| 509 |
+
"""
|
| 510 |
+
if tag is None:
|
| 511 |
+
latest_path = os.path.join(checkpoint_dir, 'latest')
|
| 512 |
+
if os.path.isfile(latest_path):
|
| 513 |
+
with open(latest_path, 'r') as fd:
|
| 514 |
+
tag = fd.read().strip()
|
| 515 |
+
else:
|
| 516 |
+
raise ValueError(f"Unable to find 'latest' file at {latest_path}")
|
| 517 |
+
|
| 518 |
+
ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
|
| 519 |
+
|
| 520 |
+
if not os.path.isdir(ds_checkpoint_dir):
|
| 521 |
+
raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
|
| 522 |
+
|
| 523 |
+
return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
|
| 524 |
+
|
| 525 |
+
|
| 526 |
+
def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
|
| 527 |
+
output_dir,
|
| 528 |
+
max_shard_size="5GB",
|
| 529 |
+
safe_serialization=False,
|
| 530 |
+
tag=None,
|
| 531 |
+
exclude_frozen_parameters=False):
|
| 532 |
+
"""
|
| 533 |
+
Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
|
| 534 |
+
loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
|
| 535 |
+
|
| 536 |
+
Args:
|
| 537 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
|
| 538 |
+
- ``output_dir``: directory to the pytorch fp32 state_dict output files
|
| 539 |
+
- ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
|
| 540 |
+
- ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
|
| 541 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
|
| 542 |
+
- ``exclude_frozen_parameters``: exclude frozen parameters
|
| 543 |
+
"""
|
| 544 |
+
# Dependency pre-check
|
| 545 |
+
if safe_serialization:
|
| 546 |
+
try:
|
| 547 |
+
from safetensors.torch import save_file
|
| 548 |
+
except ImportError:
|
| 549 |
+
print('If you want to use `safe_serialization`, please `pip install safetensors`')
|
| 550 |
+
raise
|
| 551 |
+
if max_shard_size is not None:
|
| 552 |
+
try:
|
| 553 |
+
from huggingface_hub import split_torch_state_dict_into_shards
|
| 554 |
+
except ImportError:
|
| 555 |
+
print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
|
| 556 |
+
raise
|
| 557 |
+
|
| 558 |
+
# Convert zero checkpoint to state_dict
|
| 559 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
|
| 560 |
+
|
| 561 |
+
# Shard the model if it is too big.
|
| 562 |
+
weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
|
| 563 |
+
if max_shard_size is not None:
|
| 564 |
+
filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
|
| 565 |
+
state_dict_split = split_torch_state_dict_into_shards(state_dict,
|
| 566 |
+
filename_pattern=filename_pattern,
|
| 567 |
+
max_shard_size=max_shard_size)
|
| 568 |
+
else:
|
| 569 |
+
from collections import namedtuple
|
| 570 |
+
StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
|
| 571 |
+
state_dict_split = StateDictSplit(is_sharded=False,
|
| 572 |
+
filename_to_tensors={weights_name: list(state_dict.keys())})
|
| 573 |
+
|
| 574 |
+
# Save the model
|
| 575 |
+
filename_to_tensors = state_dict_split.filename_to_tensors.items()
|
| 576 |
+
for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
|
| 577 |
+
shard = {tensor: state_dict[tensor].contiguous() for tensor in tensors}
|
| 578 |
+
output_path = os.path.join(output_dir, shard_file)
|
| 579 |
+
if safe_serialization:
|
| 580 |
+
save_file(shard, output_path, metadata={"format": "pt"})
|
| 581 |
+
else:
|
| 582 |
+
torch.save(shard, output_path)
|
| 583 |
+
|
| 584 |
+
# Save index if sharded
|
| 585 |
+
if state_dict_split.is_sharded:
|
| 586 |
+
index = {
|
| 587 |
+
"metadata": state_dict_split.metadata,
|
| 588 |
+
"weight_map": state_dict_split.tensor_to_filename,
|
| 589 |
+
}
|
| 590 |
+
save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
|
| 591 |
+
save_index_file = os.path.join(output_dir, save_index_file)
|
| 592 |
+
with open(save_index_file, "w", encoding="utf-8") as f:
|
| 593 |
+
content = json.dumps(index, indent=2, sort_keys=True) + "\n"
|
| 594 |
+
f.write(content)
|
| 595 |
+
|
| 596 |
+
|
| 597 |
+
def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
|
| 598 |
+
"""
|
| 599 |
+
1. Put the provided model to cpu
|
| 600 |
+
2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
|
| 601 |
+
3. Load it into the provided model
|
| 602 |
+
|
| 603 |
+
Args:
|
| 604 |
+
- ``model``: the model object to update
|
| 605 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
|
| 606 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
|
| 607 |
+
|
| 608 |
+
Returns:
|
| 609 |
+
- ``model`: modified model
|
| 610 |
+
|
| 611 |
+
Make sure you have plenty of CPU memory available before you call this function. If you don't
|
| 612 |
+
have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
|
| 613 |
+
conveniently placed for you in the checkpoint folder.
|
| 614 |
+
|
| 615 |
+
A typical usage might be ::
|
| 616 |
+
|
| 617 |
+
from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
|
| 618 |
+
model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
|
| 619 |
+
# submit to model hub or save the model to share with others
|
| 620 |
+
|
| 621 |
+
Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
|
| 622 |
+
of the same application. i.e. you will need to re-initialize the deepspeed engine, since
|
| 623 |
+
``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
|
| 624 |
+
|
| 625 |
+
"""
|
| 626 |
+
logger.info(f"Extracting fp32 weights")
|
| 627 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
|
| 628 |
+
|
| 629 |
+
logger.info(f"Overwriting model with fp32 weights")
|
| 630 |
+
model = model.cpu()
|
| 631 |
+
model.load_state_dict(state_dict, strict=False)
|
| 632 |
+
|
| 633 |
+
return model
|
| 634 |
+
|
| 635 |
+
|
| 636 |
+
if __name__ == "__main__":
|
| 637 |
+
parser = argparse.ArgumentParser()
|
| 638 |
+
parser.add_argument("checkpoint_dir",
|
| 639 |
+
type=str,
|
| 640 |
+
help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
|
| 641 |
+
parser.add_argument("output_dir",
|
| 642 |
+
type=str,
|
| 643 |
+
help="directory to the pytorch fp32 state_dict output files"
|
| 644 |
+
"(e.g. path/checkpoint-12-output/)")
|
| 645 |
+
parser.add_argument(
|
| 646 |
+
"--max_shard_size",
|
| 647 |
+
type=str,
|
| 648 |
+
default="5GB",
|
| 649 |
+
help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
|
| 650 |
+
"lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
|
| 651 |
+
"We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
|
| 652 |
+
"without CPU OOM issues.")
|
| 653 |
+
parser.add_argument(
|
| 654 |
+
"--safe_serialization",
|
| 655 |
+
default=False,
|
| 656 |
+
action='store_true',
|
| 657 |
+
help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
|
| 658 |
+
parser.add_argument("-t",
|
| 659 |
+
"--tag",
|
| 660 |
+
type=str,
|
| 661 |
+
default=None,
|
| 662 |
+
help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
|
| 663 |
+
parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
|
| 664 |
+
parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
|
| 665 |
+
args = parser.parse_args()
|
| 666 |
+
|
| 667 |
+
debug = args.debug
|
| 668 |
+
|
| 669 |
+
convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
|
| 670 |
+
args.output_dir,
|
| 671 |
+
max_shard_size=args.max_shard_size,
|
| 672 |
+
safe_serialization=args.safe_serialization,
|
| 673 |
+
tag=args.tag,
|
| 674 |
+
exclude_frozen_parameters=args.exclude_frozen_parameters)
|