Upload 7 files
Browse files- adapter_config.json +34 -0
- adapter_model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1447 -0
- training_args.bin +3 -0
adapter_config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "/media/houfengzhen/\u5f20\u7490\u7490/jiangchenlu/Qwen2.5-Coder-7B-Instruct",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layer_replication": null,
|
| 10 |
+
"layers_pattern": null,
|
| 11 |
+
"layers_to_transform": null,
|
| 12 |
+
"loftq_config": {},
|
| 13 |
+
"lora_alpha": 16,
|
| 14 |
+
"lora_dropout": 0.1,
|
| 15 |
+
"megatron_config": null,
|
| 16 |
+
"megatron_core": "megatron.core",
|
| 17 |
+
"modules_to_save": null,
|
| 18 |
+
"peft_type": "LORA",
|
| 19 |
+
"r": 64,
|
| 20 |
+
"rank_pattern": {},
|
| 21 |
+
"revision": null,
|
| 22 |
+
"target_modules": [
|
| 23 |
+
"gate_proj",
|
| 24 |
+
"k_proj",
|
| 25 |
+
"down_proj",
|
| 26 |
+
"q_proj",
|
| 27 |
+
"v_proj",
|
| 28 |
+
"o_proj",
|
| 29 |
+
"up_proj"
|
| 30 |
+
],
|
| 31 |
+
"task_type": "CAUSAL_LM",
|
| 32 |
+
"use_dora": false,
|
| 33 |
+
"use_rslora": false
|
| 34 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bfc4a3e561e98da6e8731e34daa1c5c01d646540ec7926bf5fe89eb5842fe350
|
| 3 |
+
size 323014560
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:544fccd3600586965c82589629039ad424250ea33cd70da3a73b2a873f7b3577
|
| 3 |
+
size 646250674
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:49781f7c3a945e40298a10d6457930d17b61001c01fbf1a6cd1660c7d4fe301b
|
| 3 |
+
size 14168
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8dd8ff83532384ff919fb02b232d56741671d82b30885e230925513f436a994f
|
| 3 |
+
size 1056
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1447 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 20205,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.014847809948032665,
|
| 13 |
+
"grad_norm": 0.14453125,
|
| 14 |
+
"learning_rate": 9.950507300173225e-05,
|
| 15 |
+
"loss": 0.3437,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02969561989606533,
|
| 20 |
+
"grad_norm": 0.10986328125,
|
| 21 |
+
"learning_rate": 9.90101460034645e-05,
|
| 22 |
+
"loss": 0.305,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.044543429844097995,
|
| 27 |
+
"grad_norm": 0.1630859375,
|
| 28 |
+
"learning_rate": 9.851521900519674e-05,
|
| 29 |
+
"loss": 0.2885,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.05939123979213066,
|
| 34 |
+
"grad_norm": 0.162109375,
|
| 35 |
+
"learning_rate": 9.802029200692899e-05,
|
| 36 |
+
"loss": 0.2781,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.07423904974016332,
|
| 41 |
+
"grad_norm": 0.193359375,
|
| 42 |
+
"learning_rate": 9.752536500866123e-05,
|
| 43 |
+
"loss": 0.2746,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.08908685968819599,
|
| 48 |
+
"grad_norm": 0.1357421875,
|
| 49 |
+
"learning_rate": 9.703043801039347e-05,
|
| 50 |
+
"loss": 0.2658,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.10393466963622866,
|
| 55 |
+
"grad_norm": 0.1689453125,
|
| 56 |
+
"learning_rate": 9.653551101212572e-05,
|
| 57 |
+
"loss": 0.2709,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.11878247958426132,
|
| 62 |
+
"grad_norm": 0.17578125,
|
| 63 |
+
"learning_rate": 9.604058401385795e-05,
|
| 64 |
+
"loss": 0.2634,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.133630289532294,
|
| 69 |
+
"grad_norm": 0.15234375,
|
| 70 |
+
"learning_rate": 9.55456570155902e-05,
|
| 71 |
+
"loss": 0.2688,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.14847809948032664,
|
| 76 |
+
"grad_norm": 0.146484375,
|
| 77 |
+
"learning_rate": 9.505073001732245e-05,
|
| 78 |
+
"loss": 0.261,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1633259094283593,
|
| 83 |
+
"grad_norm": 0.12158203125,
|
| 84 |
+
"learning_rate": 9.455580301905469e-05,
|
| 85 |
+
"loss": 0.2537,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.17817371937639198,
|
| 90 |
+
"grad_norm": 0.1845703125,
|
| 91 |
+
"learning_rate": 9.406087602078693e-05,
|
| 92 |
+
"loss": 0.2592,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.19302152932442465,
|
| 97 |
+
"grad_norm": 0.234375,
|
| 98 |
+
"learning_rate": 9.356594902251918e-05,
|
| 99 |
+
"loss": 0.2529,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.20786933927245732,
|
| 104 |
+
"grad_norm": 0.1064453125,
|
| 105 |
+
"learning_rate": 9.307102202425143e-05,
|
| 106 |
+
"loss": 0.2489,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.22271714922049,
|
| 111 |
+
"grad_norm": 0.166015625,
|
| 112 |
+
"learning_rate": 9.257609502598367e-05,
|
| 113 |
+
"loss": 0.2539,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.23756495916852263,
|
| 118 |
+
"grad_norm": 0.1669921875,
|
| 119 |
+
"learning_rate": 9.208116802771592e-05,
|
| 120 |
+
"loss": 0.2449,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.2524127691165553,
|
| 125 |
+
"grad_norm": 0.2109375,
|
| 126 |
+
"learning_rate": 9.158624102944816e-05,
|
| 127 |
+
"loss": 0.2457,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.267260579064588,
|
| 132 |
+
"grad_norm": 0.1904296875,
|
| 133 |
+
"learning_rate": 9.109131403118041e-05,
|
| 134 |
+
"loss": 0.2424,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.28210838901262064,
|
| 139 |
+
"grad_norm": 0.23828125,
|
| 140 |
+
"learning_rate": 9.059638703291266e-05,
|
| 141 |
+
"loss": 0.2485,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2969561989606533,
|
| 146 |
+
"grad_norm": 0.203125,
|
| 147 |
+
"learning_rate": 9.01014600346449e-05,
|
| 148 |
+
"loss": 0.2534,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.311804008908686,
|
| 153 |
+
"grad_norm": 0.17578125,
|
| 154 |
+
"learning_rate": 8.960653303637714e-05,
|
| 155 |
+
"loss": 0.238,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.3266518188567186,
|
| 160 |
+
"grad_norm": 0.158203125,
|
| 161 |
+
"learning_rate": 8.911160603810938e-05,
|
| 162 |
+
"loss": 0.2409,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.3414996288047513,
|
| 167 |
+
"grad_norm": 0.177734375,
|
| 168 |
+
"learning_rate": 8.861667903984163e-05,
|
| 169 |
+
"loss": 0.2484,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.35634743875278396,
|
| 174 |
+
"grad_norm": 0.1689453125,
|
| 175 |
+
"learning_rate": 8.812175204157386e-05,
|
| 176 |
+
"loss": 0.2401,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.3711952487008166,
|
| 181 |
+
"grad_norm": 0.15625,
|
| 182 |
+
"learning_rate": 8.762682504330611e-05,
|
| 183 |
+
"loss": 0.2372,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.3860430586488493,
|
| 188 |
+
"grad_norm": 0.193359375,
|
| 189 |
+
"learning_rate": 8.713189804503836e-05,
|
| 190 |
+
"loss": 0.2346,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.40089086859688194,
|
| 195 |
+
"grad_norm": 0.19140625,
|
| 196 |
+
"learning_rate": 8.663697104677061e-05,
|
| 197 |
+
"loss": 0.2307,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.41573867854491464,
|
| 202 |
+
"grad_norm": 0.15234375,
|
| 203 |
+
"learning_rate": 8.614204404850284e-05,
|
| 204 |
+
"loss": 0.2276,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.4305864884929473,
|
| 209 |
+
"grad_norm": 0.1513671875,
|
| 210 |
+
"learning_rate": 8.56471170502351e-05,
|
| 211 |
+
"loss": 0.2386,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.44543429844098,
|
| 216 |
+
"grad_norm": 0.189453125,
|
| 217 |
+
"learning_rate": 8.515219005196734e-05,
|
| 218 |
+
"loss": 0.2313,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.4602821083890126,
|
| 223 |
+
"grad_norm": 0.208984375,
|
| 224 |
+
"learning_rate": 8.465726305369959e-05,
|
| 225 |
+
"loss": 0.245,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.47512991833704527,
|
| 230 |
+
"grad_norm": 0.1943359375,
|
| 231 |
+
"learning_rate": 8.416233605543183e-05,
|
| 232 |
+
"loss": 0.2278,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.48997772828507796,
|
| 237 |
+
"grad_norm": 0.21484375,
|
| 238 |
+
"learning_rate": 8.366740905716407e-05,
|
| 239 |
+
"loss": 0.2346,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5048255382331106,
|
| 244 |
+
"grad_norm": 0.173828125,
|
| 245 |
+
"learning_rate": 8.317248205889632e-05,
|
| 246 |
+
"loss": 0.2364,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5196733481811433,
|
| 251 |
+
"grad_norm": 0.1611328125,
|
| 252 |
+
"learning_rate": 8.267755506062856e-05,
|
| 253 |
+
"loss": 0.2238,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.534521158129176,
|
| 258 |
+
"grad_norm": 0.154296875,
|
| 259 |
+
"learning_rate": 8.21826280623608e-05,
|
| 260 |
+
"loss": 0.2396,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.5493689680772086,
|
| 265 |
+
"grad_norm": 0.2255859375,
|
| 266 |
+
"learning_rate": 8.168770106409304e-05,
|
| 267 |
+
"loss": 0.2254,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.5642167780252413,
|
| 272 |
+
"grad_norm": 0.18359375,
|
| 273 |
+
"learning_rate": 8.119277406582529e-05,
|
| 274 |
+
"loss": 0.2259,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.579064587973274,
|
| 279 |
+
"grad_norm": 0.158203125,
|
| 280 |
+
"learning_rate": 8.069784706755754e-05,
|
| 281 |
+
"loss": 0.2273,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.5939123979213066,
|
| 286 |
+
"grad_norm": 0.1669921875,
|
| 287 |
+
"learning_rate": 8.020292006928979e-05,
|
| 288 |
+
"loss": 0.2274,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.6087602078693393,
|
| 293 |
+
"grad_norm": 0.166015625,
|
| 294 |
+
"learning_rate": 7.970799307102202e-05,
|
| 295 |
+
"loss": 0.2322,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.623608017817372,
|
| 300 |
+
"grad_norm": 0.2138671875,
|
| 301 |
+
"learning_rate": 7.921306607275427e-05,
|
| 302 |
+
"loss": 0.2274,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.6384558277654045,
|
| 307 |
+
"grad_norm": 0.1416015625,
|
| 308 |
+
"learning_rate": 7.871813907448652e-05,
|
| 309 |
+
"loss": 0.2274,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.6533036377134372,
|
| 314 |
+
"grad_norm": 0.2060546875,
|
| 315 |
+
"learning_rate": 7.822321207621877e-05,
|
| 316 |
+
"loss": 0.2362,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.6681514476614699,
|
| 321 |
+
"grad_norm": 0.1572265625,
|
| 322 |
+
"learning_rate": 7.7728285077951e-05,
|
| 323 |
+
"loss": 0.2367,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.6829992576095026,
|
| 328 |
+
"grad_norm": 0.1650390625,
|
| 329 |
+
"learning_rate": 7.723335807968325e-05,
|
| 330 |
+
"loss": 0.2287,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.6978470675575352,
|
| 335 |
+
"grad_norm": 0.1259765625,
|
| 336 |
+
"learning_rate": 7.67384310814155e-05,
|
| 337 |
+
"loss": 0.233,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.7126948775055679,
|
| 342 |
+
"grad_norm": 0.1455078125,
|
| 343 |
+
"learning_rate": 7.624350408314775e-05,
|
| 344 |
+
"loss": 0.2243,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.7275426874536006,
|
| 349 |
+
"grad_norm": 0.14453125,
|
| 350 |
+
"learning_rate": 7.574857708487998e-05,
|
| 351 |
+
"loss": 0.2311,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.7423904974016332,
|
| 356 |
+
"grad_norm": 0.251953125,
|
| 357 |
+
"learning_rate": 7.525365008661222e-05,
|
| 358 |
+
"loss": 0.229,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.7572383073496659,
|
| 363 |
+
"grad_norm": 0.234375,
|
| 364 |
+
"learning_rate": 7.475872308834447e-05,
|
| 365 |
+
"loss": 0.2275,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.7720861172976986,
|
| 370 |
+
"grad_norm": 0.28125,
|
| 371 |
+
"learning_rate": 7.426379609007672e-05,
|
| 372 |
+
"loss": 0.2235,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.7869339272457313,
|
| 377 |
+
"grad_norm": 0.2431640625,
|
| 378 |
+
"learning_rate": 7.376886909180895e-05,
|
| 379 |
+
"loss": 0.2234,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.8017817371937639,
|
| 384 |
+
"grad_norm": 0.2119140625,
|
| 385 |
+
"learning_rate": 7.32739420935412e-05,
|
| 386 |
+
"loss": 0.2236,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.8166295471417966,
|
| 391 |
+
"grad_norm": 0.212890625,
|
| 392 |
+
"learning_rate": 7.277901509527345e-05,
|
| 393 |
+
"loss": 0.2265,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.8314773570898293,
|
| 398 |
+
"grad_norm": 0.236328125,
|
| 399 |
+
"learning_rate": 7.22840880970057e-05,
|
| 400 |
+
"loss": 0.2121,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.8463251670378619,
|
| 405 |
+
"grad_norm": 0.1953125,
|
| 406 |
+
"learning_rate": 7.178916109873795e-05,
|
| 407 |
+
"loss": 0.2225,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.8611729769858946,
|
| 412 |
+
"grad_norm": 0.2119140625,
|
| 413 |
+
"learning_rate": 7.129423410047018e-05,
|
| 414 |
+
"loss": 0.2196,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.8760207869339273,
|
| 419 |
+
"grad_norm": 0.171875,
|
| 420 |
+
"learning_rate": 7.079930710220243e-05,
|
| 421 |
+
"loss": 0.2235,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.89086859688196,
|
| 426 |
+
"grad_norm": 0.2333984375,
|
| 427 |
+
"learning_rate": 7.030438010393468e-05,
|
| 428 |
+
"loss": 0.2256,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.9057164068299925,
|
| 433 |
+
"grad_norm": 0.279296875,
|
| 434 |
+
"learning_rate": 6.980945310566693e-05,
|
| 435 |
+
"loss": 0.2217,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.9205642167780252,
|
| 440 |
+
"grad_norm": 0.16015625,
|
| 441 |
+
"learning_rate": 6.931452610739916e-05,
|
| 442 |
+
"loss": 0.223,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.9354120267260579,
|
| 447 |
+
"grad_norm": 0.287109375,
|
| 448 |
+
"learning_rate": 6.881959910913141e-05,
|
| 449 |
+
"loss": 0.2165,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.9502598366740905,
|
| 454 |
+
"grad_norm": 0.255859375,
|
| 455 |
+
"learning_rate": 6.832467211086365e-05,
|
| 456 |
+
"loss": 0.2222,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.9651076466221232,
|
| 461 |
+
"grad_norm": 0.2080078125,
|
| 462 |
+
"learning_rate": 6.78297451125959e-05,
|
| 463 |
+
"loss": 0.2243,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.9799554565701559,
|
| 468 |
+
"grad_norm": 0.291015625,
|
| 469 |
+
"learning_rate": 6.733481811432813e-05,
|
| 470 |
+
"loss": 0.2261,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.9948032665181886,
|
| 475 |
+
"grad_norm": 0.26171875,
|
| 476 |
+
"learning_rate": 6.683989111606038e-05,
|
| 477 |
+
"loss": 0.2195,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.0096510764662212,
|
| 482 |
+
"grad_norm": 0.2138671875,
|
| 483 |
+
"learning_rate": 6.634496411779263e-05,
|
| 484 |
+
"loss": 0.2005,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.024498886414254,
|
| 489 |
+
"grad_norm": 0.21875,
|
| 490 |
+
"learning_rate": 6.585003711952488e-05,
|
| 491 |
+
"loss": 0.1892,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.0393466963622866,
|
| 496 |
+
"grad_norm": 0.21484375,
|
| 497 |
+
"learning_rate": 6.535511012125711e-05,
|
| 498 |
+
"loss": 0.1842,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.0541945063103193,
|
| 503 |
+
"grad_norm": 0.1982421875,
|
| 504 |
+
"learning_rate": 6.486018312298936e-05,
|
| 505 |
+
"loss": 0.1897,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.069042316258352,
|
| 510 |
+
"grad_norm": 0.1884765625,
|
| 511 |
+
"learning_rate": 6.436525612472161e-05,
|
| 512 |
+
"loss": 0.1831,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.0838901262063845,
|
| 517 |
+
"grad_norm": 0.2578125,
|
| 518 |
+
"learning_rate": 6.387032912645386e-05,
|
| 519 |
+
"loss": 0.1891,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.0987379361544172,
|
| 524 |
+
"grad_norm": 0.17578125,
|
| 525 |
+
"learning_rate": 6.33754021281861e-05,
|
| 526 |
+
"loss": 0.1884,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.1135857461024499,
|
| 531 |
+
"grad_norm": 0.2451171875,
|
| 532 |
+
"learning_rate": 6.288047512991834e-05,
|
| 533 |
+
"loss": 0.1904,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.1284335560504826,
|
| 538 |
+
"grad_norm": 0.267578125,
|
| 539 |
+
"learning_rate": 6.238554813165059e-05,
|
| 540 |
+
"loss": 0.1899,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.1432813659985153,
|
| 545 |
+
"grad_norm": 0.30859375,
|
| 546 |
+
"learning_rate": 6.189062113338282e-05,
|
| 547 |
+
"loss": 0.1884,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.158129175946548,
|
| 552 |
+
"grad_norm": 0.255859375,
|
| 553 |
+
"learning_rate": 6.139569413511507e-05,
|
| 554 |
+
"loss": 0.1905,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.1729769858945804,
|
| 559 |
+
"grad_norm": 0.2373046875,
|
| 560 |
+
"learning_rate": 6.0900767136847315e-05,
|
| 561 |
+
"loss": 0.183,
|
| 562 |
+
"step": 7900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.1878247958426131,
|
| 566 |
+
"grad_norm": 0.2333984375,
|
| 567 |
+
"learning_rate": 6.0405840138579564e-05,
|
| 568 |
+
"loss": 0.1841,
|
| 569 |
+
"step": 8000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.2026726057906458,
|
| 573 |
+
"grad_norm": 0.2490234375,
|
| 574 |
+
"learning_rate": 5.9910913140311805e-05,
|
| 575 |
+
"loss": 0.1873,
|
| 576 |
+
"step": 8100
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.2175204157386785,
|
| 580 |
+
"grad_norm": 0.267578125,
|
| 581 |
+
"learning_rate": 5.9415986142044054e-05,
|
| 582 |
+
"loss": 0.1817,
|
| 583 |
+
"step": 8200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.2323682256867112,
|
| 587 |
+
"grad_norm": 0.28515625,
|
| 588 |
+
"learning_rate": 5.892105914377629e-05,
|
| 589 |
+
"loss": 0.1824,
|
| 590 |
+
"step": 8300
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.247216035634744,
|
| 594 |
+
"grad_norm": 0.26953125,
|
| 595 |
+
"learning_rate": 5.842613214550854e-05,
|
| 596 |
+
"loss": 0.19,
|
| 597 |
+
"step": 8400
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.2620638455827766,
|
| 601 |
+
"grad_norm": 0.236328125,
|
| 602 |
+
"learning_rate": 5.7931205147240786e-05,
|
| 603 |
+
"loss": 0.1859,
|
| 604 |
+
"step": 8500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.2769116555308093,
|
| 608 |
+
"grad_norm": 0.255859375,
|
| 609 |
+
"learning_rate": 5.7436278148973035e-05,
|
| 610 |
+
"loss": 0.1877,
|
| 611 |
+
"step": 8600
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.2917594654788418,
|
| 615 |
+
"grad_norm": 0.1826171875,
|
| 616 |
+
"learning_rate": 5.694135115070527e-05,
|
| 617 |
+
"loss": 0.1813,
|
| 618 |
+
"step": 8700
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.3066072754268745,
|
| 622 |
+
"grad_norm": 0.365234375,
|
| 623 |
+
"learning_rate": 5.644642415243752e-05,
|
| 624 |
+
"loss": 0.186,
|
| 625 |
+
"step": 8800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.3214550853749072,
|
| 629 |
+
"grad_norm": 0.2041015625,
|
| 630 |
+
"learning_rate": 5.595149715416976e-05,
|
| 631 |
+
"loss": 0.1861,
|
| 632 |
+
"step": 8900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.3363028953229399,
|
| 636 |
+
"grad_norm": 0.248046875,
|
| 637 |
+
"learning_rate": 5.545657015590201e-05,
|
| 638 |
+
"loss": 0.1858,
|
| 639 |
+
"step": 9000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.3511507052709726,
|
| 643 |
+
"grad_norm": 0.2021484375,
|
| 644 |
+
"learning_rate": 5.4961643157634244e-05,
|
| 645 |
+
"loss": 0.1827,
|
| 646 |
+
"step": 9100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.365998515219005,
|
| 650 |
+
"grad_norm": 0.2080078125,
|
| 651 |
+
"learning_rate": 5.446671615936649e-05,
|
| 652 |
+
"loss": 0.1778,
|
| 653 |
+
"step": 9200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.3808463251670378,
|
| 657 |
+
"grad_norm": 0.26171875,
|
| 658 |
+
"learning_rate": 5.397178916109874e-05,
|
| 659 |
+
"loss": 0.1891,
|
| 660 |
+
"step": 9300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.3956941351150705,
|
| 664 |
+
"grad_norm": 0.294921875,
|
| 665 |
+
"learning_rate": 5.347686216283099e-05,
|
| 666 |
+
"loss": 0.1842,
|
| 667 |
+
"step": 9400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.4105419450631032,
|
| 671 |
+
"grad_norm": 0.2392578125,
|
| 672 |
+
"learning_rate": 5.298193516456323e-05,
|
| 673 |
+
"loss": 0.187,
|
| 674 |
+
"step": 9500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.4253897550111359,
|
| 678 |
+
"grad_norm": 0.2216796875,
|
| 679 |
+
"learning_rate": 5.248700816629547e-05,
|
| 680 |
+
"loss": 0.1897,
|
| 681 |
+
"step": 9600
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.4402375649591685,
|
| 685 |
+
"grad_norm": 0.322265625,
|
| 686 |
+
"learning_rate": 5.1992081168027716e-05,
|
| 687 |
+
"loss": 0.1873,
|
| 688 |
+
"step": 9700
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.4550853749072012,
|
| 692 |
+
"grad_norm": 0.291015625,
|
| 693 |
+
"learning_rate": 5.1497154169759965e-05,
|
| 694 |
+
"loss": 0.188,
|
| 695 |
+
"step": 9800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.469933184855234,
|
| 699 |
+
"grad_norm": 0.2119140625,
|
| 700 |
+
"learning_rate": 5.100222717149221e-05,
|
| 701 |
+
"loss": 0.1824,
|
| 702 |
+
"step": 9900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.4847809948032666,
|
| 706 |
+
"grad_norm": 0.185546875,
|
| 707 |
+
"learning_rate": 5.050730017322445e-05,
|
| 708 |
+
"loss": 0.1827,
|
| 709 |
+
"step": 10000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.4996288047512991,
|
| 713 |
+
"grad_norm": 0.33203125,
|
| 714 |
+
"learning_rate": 5.00123731749567e-05,
|
| 715 |
+
"loss": 0.185,
|
| 716 |
+
"step": 10100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.5144766146993318,
|
| 720 |
+
"grad_norm": 0.1884765625,
|
| 721 |
+
"learning_rate": 4.951744617668894e-05,
|
| 722 |
+
"loss": 0.1813,
|
| 723 |
+
"step": 10200
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.5293244246473645,
|
| 727 |
+
"grad_norm": 0.2294921875,
|
| 728 |
+
"learning_rate": 4.902251917842118e-05,
|
| 729 |
+
"loss": 0.1796,
|
| 730 |
+
"step": 10300
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.5441722345953972,
|
| 734 |
+
"grad_norm": 0.296875,
|
| 735 |
+
"learning_rate": 4.852759218015343e-05,
|
| 736 |
+
"loss": 0.1887,
|
| 737 |
+
"step": 10400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.5590200445434297,
|
| 741 |
+
"grad_norm": 0.166015625,
|
| 742 |
+
"learning_rate": 4.803266518188567e-05,
|
| 743 |
+
"loss": 0.1845,
|
| 744 |
+
"step": 10500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.5738678544914624,
|
| 748 |
+
"grad_norm": 0.326171875,
|
| 749 |
+
"learning_rate": 4.753773818361792e-05,
|
| 750 |
+
"loss": 0.1784,
|
| 751 |
+
"step": 10600
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.588715664439495,
|
| 755 |
+
"grad_norm": 0.1982421875,
|
| 756 |
+
"learning_rate": 4.704281118535016e-05,
|
| 757 |
+
"loss": 0.1842,
|
| 758 |
+
"step": 10700
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.6035634743875278,
|
| 762 |
+
"grad_norm": 0.314453125,
|
| 763 |
+
"learning_rate": 4.654788418708241e-05,
|
| 764 |
+
"loss": 0.188,
|
| 765 |
+
"step": 10800
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.6184112843355605,
|
| 769 |
+
"grad_norm": 0.25,
|
| 770 |
+
"learning_rate": 4.605295718881465e-05,
|
| 771 |
+
"loss": 0.1844,
|
| 772 |
+
"step": 10900
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.6332590942835932,
|
| 776 |
+
"grad_norm": 0.380859375,
|
| 777 |
+
"learning_rate": 4.5558030190546894e-05,
|
| 778 |
+
"loss": 0.1802,
|
| 779 |
+
"step": 11000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.6481069042316259,
|
| 783 |
+
"grad_norm": 0.2314453125,
|
| 784 |
+
"learning_rate": 4.506310319227914e-05,
|
| 785 |
+
"loss": 0.1881,
|
| 786 |
+
"step": 11100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.6629547141796586,
|
| 790 |
+
"grad_norm": 0.37109375,
|
| 791 |
+
"learning_rate": 4.4568176194011384e-05,
|
| 792 |
+
"loss": 0.185,
|
| 793 |
+
"step": 11200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.6778025241276913,
|
| 797 |
+
"grad_norm": 0.283203125,
|
| 798 |
+
"learning_rate": 4.407324919574363e-05,
|
| 799 |
+
"loss": 0.1818,
|
| 800 |
+
"step": 11300
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.692650334075724,
|
| 804 |
+
"grad_norm": 0.25,
|
| 805 |
+
"learning_rate": 4.3578322197475875e-05,
|
| 806 |
+
"loss": 0.182,
|
| 807 |
+
"step": 11400
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.7074981440237567,
|
| 811 |
+
"grad_norm": 0.25390625,
|
| 812 |
+
"learning_rate": 4.3083395199208124e-05,
|
| 813 |
+
"loss": 0.184,
|
| 814 |
+
"step": 11500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.7223459539717891,
|
| 818 |
+
"grad_norm": 0.28515625,
|
| 819 |
+
"learning_rate": 4.2588468200940365e-05,
|
| 820 |
+
"loss": 0.1923,
|
| 821 |
+
"step": 11600
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.7371937639198218,
|
| 825 |
+
"grad_norm": 0.2236328125,
|
| 826 |
+
"learning_rate": 4.209354120267261e-05,
|
| 827 |
+
"loss": 0.1812,
|
| 828 |
+
"step": 11700
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.7520415738678545,
|
| 832 |
+
"grad_norm": 0.1865234375,
|
| 833 |
+
"learning_rate": 4.159861420440485e-05,
|
| 834 |
+
"loss": 0.1756,
|
| 835 |
+
"step": 11800
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.766889383815887,
|
| 839 |
+
"grad_norm": 0.31640625,
|
| 840 |
+
"learning_rate": 4.11036872061371e-05,
|
| 841 |
+
"loss": 0.1878,
|
| 842 |
+
"step": 11900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.7817371937639197,
|
| 846 |
+
"grad_norm": 0.32421875,
|
| 847 |
+
"learning_rate": 4.060876020786934e-05,
|
| 848 |
+
"loss": 0.18,
|
| 849 |
+
"step": 12000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.7965850037119524,
|
| 853 |
+
"grad_norm": 0.265625,
|
| 854 |
+
"learning_rate": 4.011383320960159e-05,
|
| 855 |
+
"loss": 0.1802,
|
| 856 |
+
"step": 12100
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.811432813659985,
|
| 860 |
+
"grad_norm": 0.234375,
|
| 861 |
+
"learning_rate": 3.961890621133383e-05,
|
| 862 |
+
"loss": 0.1802,
|
| 863 |
+
"step": 12200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.8262806236080178,
|
| 867 |
+
"grad_norm": 0.31640625,
|
| 868 |
+
"learning_rate": 3.912397921306608e-05,
|
| 869 |
+
"loss": 0.1842,
|
| 870 |
+
"step": 12300
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.8411284335560505,
|
| 874 |
+
"grad_norm": 0.2109375,
|
| 875 |
+
"learning_rate": 3.8629052214798314e-05,
|
| 876 |
+
"loss": 0.1855,
|
| 877 |
+
"step": 12400
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.8559762435040832,
|
| 881 |
+
"grad_norm": 0.271484375,
|
| 882 |
+
"learning_rate": 3.813412521653056e-05,
|
| 883 |
+
"loss": 0.1846,
|
| 884 |
+
"step": 12500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.8708240534521159,
|
| 888 |
+
"grad_norm": 0.275390625,
|
| 889 |
+
"learning_rate": 3.7639198218262804e-05,
|
| 890 |
+
"loss": 0.1829,
|
| 891 |
+
"step": 12600
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.8856718634001486,
|
| 895 |
+
"grad_norm": 0.2734375,
|
| 896 |
+
"learning_rate": 3.714427121999505e-05,
|
| 897 |
+
"loss": 0.1845,
|
| 898 |
+
"step": 12700
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.9005196733481813,
|
| 902 |
+
"grad_norm": 0.28125,
|
| 903 |
+
"learning_rate": 3.66493442217273e-05,
|
| 904 |
+
"loss": 0.1849,
|
| 905 |
+
"step": 12800
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.9153674832962138,
|
| 909 |
+
"grad_norm": 0.2353515625,
|
| 910 |
+
"learning_rate": 3.6154417223459543e-05,
|
| 911 |
+
"loss": 0.1849,
|
| 912 |
+
"step": 12900
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.9302152932442465,
|
| 916 |
+
"grad_norm": 0.33203125,
|
| 917 |
+
"learning_rate": 3.5659490225191785e-05,
|
| 918 |
+
"loss": 0.1782,
|
| 919 |
+
"step": 13000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.9450631031922792,
|
| 923 |
+
"grad_norm": 0.29296875,
|
| 924 |
+
"learning_rate": 3.516456322692403e-05,
|
| 925 |
+
"loss": 0.1871,
|
| 926 |
+
"step": 13100
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.9599109131403119,
|
| 930 |
+
"grad_norm": 0.23828125,
|
| 931 |
+
"learning_rate": 3.4669636228656276e-05,
|
| 932 |
+
"loss": 0.1775,
|
| 933 |
+
"step": 13200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.9747587230883443,
|
| 937 |
+
"grad_norm": 0.30078125,
|
| 938 |
+
"learning_rate": 3.417470923038852e-05,
|
| 939 |
+
"loss": 0.1815,
|
| 940 |
+
"step": 13300
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.989606533036377,
|
| 944 |
+
"grad_norm": 0.30078125,
|
| 945 |
+
"learning_rate": 3.3679782232120766e-05,
|
| 946 |
+
"loss": 0.1831,
|
| 947 |
+
"step": 13400
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.0044543429844097,
|
| 951 |
+
"grad_norm": 0.2373046875,
|
| 952 |
+
"learning_rate": 3.318485523385301e-05,
|
| 953 |
+
"loss": 0.1752,
|
| 954 |
+
"step": 13500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.0193021529324424,
|
| 958 |
+
"grad_norm": 0.298828125,
|
| 959 |
+
"learning_rate": 3.268992823558526e-05,
|
| 960 |
+
"loss": 0.172,
|
| 961 |
+
"step": 13600
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.034149962880475,
|
| 965 |
+
"grad_norm": 0.263671875,
|
| 966 |
+
"learning_rate": 3.21950012373175e-05,
|
| 967 |
+
"loss": 0.1695,
|
| 968 |
+
"step": 13700
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.048997772828508,
|
| 972 |
+
"grad_norm": 0.255859375,
|
| 973 |
+
"learning_rate": 3.170007423904974e-05,
|
| 974 |
+
"loss": 0.169,
|
| 975 |
+
"step": 13800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.0638455827765405,
|
| 979 |
+
"grad_norm": 0.2333984375,
|
| 980 |
+
"learning_rate": 3.120514724078198e-05,
|
| 981 |
+
"loss": 0.1688,
|
| 982 |
+
"step": 13900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.078693392724573,
|
| 986 |
+
"grad_norm": 0.224609375,
|
| 987 |
+
"learning_rate": 3.071022024251423e-05,
|
| 988 |
+
"loss": 0.1696,
|
| 989 |
+
"step": 14000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.093541202672606,
|
| 993 |
+
"grad_norm": 0.34375,
|
| 994 |
+
"learning_rate": 3.0215293244246473e-05,
|
| 995 |
+
"loss": 0.1683,
|
| 996 |
+
"step": 14100
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.1083890126206386,
|
| 1000 |
+
"grad_norm": 0.365234375,
|
| 1001 |
+
"learning_rate": 2.972036624597872e-05,
|
| 1002 |
+
"loss": 0.1647,
|
| 1003 |
+
"step": 14200
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.1232368225686713,
|
| 1007 |
+
"grad_norm": 0.310546875,
|
| 1008 |
+
"learning_rate": 2.9225439247710963e-05,
|
| 1009 |
+
"loss": 0.1698,
|
| 1010 |
+
"step": 14300
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.138084632516704,
|
| 1014 |
+
"grad_norm": 0.20703125,
|
| 1015 |
+
"learning_rate": 2.873051224944321e-05,
|
| 1016 |
+
"loss": 0.1648,
|
| 1017 |
+
"step": 14400
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.1529324424647363,
|
| 1021 |
+
"grad_norm": 0.224609375,
|
| 1022 |
+
"learning_rate": 2.823558525117545e-05,
|
| 1023 |
+
"loss": 0.1676,
|
| 1024 |
+
"step": 14500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.167780252412769,
|
| 1028 |
+
"grad_norm": 0.283203125,
|
| 1029 |
+
"learning_rate": 2.77406582529077e-05,
|
| 1030 |
+
"loss": 0.1676,
|
| 1031 |
+
"step": 14600
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.1826280623608016,
|
| 1035 |
+
"grad_norm": 0.240234375,
|
| 1036 |
+
"learning_rate": 2.7245731254639944e-05,
|
| 1037 |
+
"loss": 0.168,
|
| 1038 |
+
"step": 14700
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.1974758723088343,
|
| 1042 |
+
"grad_norm": 0.2734375,
|
| 1043 |
+
"learning_rate": 2.6750804256372186e-05,
|
| 1044 |
+
"loss": 0.1662,
|
| 1045 |
+
"step": 14800
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.212323682256867,
|
| 1049 |
+
"grad_norm": 0.2119140625,
|
| 1050 |
+
"learning_rate": 2.6255877258104435e-05,
|
| 1051 |
+
"loss": 0.163,
|
| 1052 |
+
"step": 14900
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.2271714922048997,
|
| 1056 |
+
"grad_norm": 0.32421875,
|
| 1057 |
+
"learning_rate": 2.5760950259836673e-05,
|
| 1058 |
+
"loss": 0.1658,
|
| 1059 |
+
"step": 15000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.2420193021529324,
|
| 1063 |
+
"grad_norm": 0.2734375,
|
| 1064 |
+
"learning_rate": 2.5266023261568922e-05,
|
| 1065 |
+
"loss": 0.1645,
|
| 1066 |
+
"step": 15100
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.256867112100965,
|
| 1070 |
+
"grad_norm": 0.291015625,
|
| 1071 |
+
"learning_rate": 2.4771096263301164e-05,
|
| 1072 |
+
"loss": 0.1563,
|
| 1073 |
+
"step": 15200
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.271714922048998,
|
| 1077 |
+
"grad_norm": 0.34375,
|
| 1078 |
+
"learning_rate": 2.427616926503341e-05,
|
| 1079 |
+
"loss": 0.1677,
|
| 1080 |
+
"step": 15300
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.2865627319970305,
|
| 1084 |
+
"grad_norm": 0.265625,
|
| 1085 |
+
"learning_rate": 2.378124226676565e-05,
|
| 1086 |
+
"loss": 0.1635,
|
| 1087 |
+
"step": 15400
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.3014105419450632,
|
| 1091 |
+
"grad_norm": 0.32421875,
|
| 1092 |
+
"learning_rate": 2.3286315268497896e-05,
|
| 1093 |
+
"loss": 0.1693,
|
| 1094 |
+
"step": 15500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.316258351893096,
|
| 1098 |
+
"grad_norm": 0.2470703125,
|
| 1099 |
+
"learning_rate": 2.2791388270230145e-05,
|
| 1100 |
+
"loss": 0.1682,
|
| 1101 |
+
"step": 15600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 2.3311061618411286,
|
| 1105 |
+
"grad_norm": 0.271484375,
|
| 1106 |
+
"learning_rate": 2.2296461271962387e-05,
|
| 1107 |
+
"loss": 0.1637,
|
| 1108 |
+
"step": 15700
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.345953971789161,
|
| 1112 |
+
"grad_norm": 0.337890625,
|
| 1113 |
+
"learning_rate": 2.1801534273694632e-05,
|
| 1114 |
+
"loss": 0.1702,
|
| 1115 |
+
"step": 15800
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 2.3608017817371936,
|
| 1119 |
+
"grad_norm": 0.3125,
|
| 1120 |
+
"learning_rate": 2.1306607275426877e-05,
|
| 1121 |
+
"loss": 0.1648,
|
| 1122 |
+
"step": 15900
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.3756495916852263,
|
| 1126 |
+
"grad_norm": 0.275390625,
|
| 1127 |
+
"learning_rate": 2.0811680277159122e-05,
|
| 1128 |
+
"loss": 0.1586,
|
| 1129 |
+
"step": 16000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.390497401633259,
|
| 1133 |
+
"grad_norm": 0.263671875,
|
| 1134 |
+
"learning_rate": 2.0316753278891364e-05,
|
| 1135 |
+
"loss": 0.1628,
|
| 1136 |
+
"step": 16100
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.4053452115812917,
|
| 1140 |
+
"grad_norm": 0.28125,
|
| 1141 |
+
"learning_rate": 1.982182628062361e-05,
|
| 1142 |
+
"loss": 0.1621,
|
| 1143 |
+
"step": 16200
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.4201930215293244,
|
| 1147 |
+
"grad_norm": 0.296875,
|
| 1148 |
+
"learning_rate": 1.9326899282355855e-05,
|
| 1149 |
+
"loss": 0.1628,
|
| 1150 |
+
"step": 16300
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.435040831477357,
|
| 1154 |
+
"grad_norm": 0.2412109375,
|
| 1155 |
+
"learning_rate": 1.8831972284088097e-05,
|
| 1156 |
+
"loss": 0.1642,
|
| 1157 |
+
"step": 16400
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 2.4498886414253898,
|
| 1161 |
+
"grad_norm": 0.314453125,
|
| 1162 |
+
"learning_rate": 1.8337045285820342e-05,
|
| 1163 |
+
"loss": 0.1659,
|
| 1164 |
+
"step": 16500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.4647364513734225,
|
| 1168 |
+
"grad_norm": 0.2734375,
|
| 1169 |
+
"learning_rate": 1.7842118287552587e-05,
|
| 1170 |
+
"loss": 0.1683,
|
| 1171 |
+
"step": 16600
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 2.479584261321455,
|
| 1175 |
+
"grad_norm": 0.322265625,
|
| 1176 |
+
"learning_rate": 1.7347191289284832e-05,
|
| 1177 |
+
"loss": 0.1655,
|
| 1178 |
+
"step": 16700
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.494432071269488,
|
| 1182 |
+
"grad_norm": 0.2734375,
|
| 1183 |
+
"learning_rate": 1.6852264291017074e-05,
|
| 1184 |
+
"loss": 0.1655,
|
| 1185 |
+
"step": 16800
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 2.5092798812175205,
|
| 1189 |
+
"grad_norm": 0.3046875,
|
| 1190 |
+
"learning_rate": 1.635733729274932e-05,
|
| 1191 |
+
"loss": 0.164,
|
| 1192 |
+
"step": 16900
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 2.5241276911655532,
|
| 1196 |
+
"grad_norm": 0.28125,
|
| 1197 |
+
"learning_rate": 1.5862410294481565e-05,
|
| 1198 |
+
"loss": 0.1745,
|
| 1199 |
+
"step": 17000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.538975501113586,
|
| 1203 |
+
"grad_norm": 0.2490234375,
|
| 1204 |
+
"learning_rate": 1.5367483296213807e-05,
|
| 1205 |
+
"loss": 0.169,
|
| 1206 |
+
"step": 17100
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 2.5538233110616186,
|
| 1210 |
+
"grad_norm": 0.314453125,
|
| 1211 |
+
"learning_rate": 1.4872556297946052e-05,
|
| 1212 |
+
"loss": 0.1695,
|
| 1213 |
+
"step": 17200
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 2.5686711210096513,
|
| 1217 |
+
"grad_norm": 0.306640625,
|
| 1218 |
+
"learning_rate": 1.4377629299678297e-05,
|
| 1219 |
+
"loss": 0.1657,
|
| 1220 |
+
"step": 17300
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.5835189309576836,
|
| 1224 |
+
"grad_norm": 0.302734375,
|
| 1225 |
+
"learning_rate": 1.3882702301410544e-05,
|
| 1226 |
+
"loss": 0.1615,
|
| 1227 |
+
"step": 17400
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 2.5983667409057163,
|
| 1231 |
+
"grad_norm": 0.263671875,
|
| 1232 |
+
"learning_rate": 1.3387775303142788e-05,
|
| 1233 |
+
"loss": 0.1675,
|
| 1234 |
+
"step": 17500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 2.613214550853749,
|
| 1238 |
+
"grad_norm": 0.29296875,
|
| 1239 |
+
"learning_rate": 1.2892848304875033e-05,
|
| 1240 |
+
"loss": 0.1646,
|
| 1241 |
+
"step": 17600
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 2.6280623608017817,
|
| 1245 |
+
"grad_norm": 0.197265625,
|
| 1246 |
+
"learning_rate": 1.2397921306607275e-05,
|
| 1247 |
+
"loss": 0.1631,
|
| 1248 |
+
"step": 17700
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 2.6429101707498144,
|
| 1252 |
+
"grad_norm": 0.2890625,
|
| 1253 |
+
"learning_rate": 1.190299430833952e-05,
|
| 1254 |
+
"loss": 0.1621,
|
| 1255 |
+
"step": 17800
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 2.657757980697847,
|
| 1259 |
+
"grad_norm": 0.29296875,
|
| 1260 |
+
"learning_rate": 1.1408067310071765e-05,
|
| 1261 |
+
"loss": 0.1682,
|
| 1262 |
+
"step": 17900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.6726057906458798,
|
| 1266 |
+
"grad_norm": 0.29296875,
|
| 1267 |
+
"learning_rate": 1.091314031180401e-05,
|
| 1268 |
+
"loss": 0.1664,
|
| 1269 |
+
"step": 18000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 2.6874536005939125,
|
| 1273 |
+
"grad_norm": 0.2431640625,
|
| 1274 |
+
"learning_rate": 1.0418213313536254e-05,
|
| 1275 |
+
"loss": 0.1686,
|
| 1276 |
+
"step": 18100
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 2.702301410541945,
|
| 1280 |
+
"grad_norm": 0.427734375,
|
| 1281 |
+
"learning_rate": 9.9232863152685e-06,
|
| 1282 |
+
"loss": 0.1653,
|
| 1283 |
+
"step": 18200
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 2.717149220489978,
|
| 1287 |
+
"grad_norm": 0.283203125,
|
| 1288 |
+
"learning_rate": 9.428359317000743e-06,
|
| 1289 |
+
"loss": 0.1704,
|
| 1290 |
+
"step": 18300
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 2.73199703043801,
|
| 1294 |
+
"grad_norm": 0.2314453125,
|
| 1295 |
+
"learning_rate": 8.933432318732986e-06,
|
| 1296 |
+
"loss": 0.1645,
|
| 1297 |
+
"step": 18400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 2.746844840386043,
|
| 1301 |
+
"grad_norm": 0.2041015625,
|
| 1302 |
+
"learning_rate": 8.438505320465232e-06,
|
| 1303 |
+
"loss": 0.1647,
|
| 1304 |
+
"step": 18500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.7616926503340755,
|
| 1308 |
+
"grad_norm": 0.333984375,
|
| 1309 |
+
"learning_rate": 7.943578322197475e-06,
|
| 1310 |
+
"loss": 0.1721,
|
| 1311 |
+
"step": 18600
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 2.776540460282108,
|
| 1315 |
+
"grad_norm": 0.330078125,
|
| 1316 |
+
"learning_rate": 7.44865132392972e-06,
|
| 1317 |
+
"loss": 0.1652,
|
| 1318 |
+
"step": 18700
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.791388270230141,
|
| 1322 |
+
"grad_norm": 0.333984375,
|
| 1323 |
+
"learning_rate": 6.953724325661966e-06,
|
| 1324 |
+
"loss": 0.1634,
|
| 1325 |
+
"step": 18800
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 2.8062360801781736,
|
| 1329 |
+
"grad_norm": 0.283203125,
|
| 1330 |
+
"learning_rate": 6.45879732739421e-06,
|
| 1331 |
+
"loss": 0.1668,
|
| 1332 |
+
"step": 18900
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 2.8210838901262063,
|
| 1336 |
+
"grad_norm": 0.248046875,
|
| 1337 |
+
"learning_rate": 5.9638703291264544e-06,
|
| 1338 |
+
"loss": 0.1642,
|
| 1339 |
+
"step": 19000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 2.835931700074239,
|
| 1343 |
+
"grad_norm": 0.2373046875,
|
| 1344 |
+
"learning_rate": 5.468943330858699e-06,
|
| 1345 |
+
"loss": 0.1594,
|
| 1346 |
+
"step": 19100
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.8507795100222717,
|
| 1350 |
+
"grad_norm": 0.224609375,
|
| 1351 |
+
"learning_rate": 4.974016332590943e-06,
|
| 1352 |
+
"loss": 0.1648,
|
| 1353 |
+
"step": 19200
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 2.8656273199703044,
|
| 1357 |
+
"grad_norm": 0.2353515625,
|
| 1358 |
+
"learning_rate": 4.479089334323188e-06,
|
| 1359 |
+
"loss": 0.1637,
|
| 1360 |
+
"step": 19300
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.880475129918337,
|
| 1364 |
+
"grad_norm": 0.2294921875,
|
| 1365 |
+
"learning_rate": 3.984162336055432e-06,
|
| 1366 |
+
"loss": 0.1728,
|
| 1367 |
+
"step": 19400
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 2.89532293986637,
|
| 1371 |
+
"grad_norm": 0.25390625,
|
| 1372 |
+
"learning_rate": 3.489235337787677e-06,
|
| 1373 |
+
"loss": 0.1646,
|
| 1374 |
+
"step": 19500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.9101707498144025,
|
| 1378 |
+
"grad_norm": 0.25,
|
| 1379 |
+
"learning_rate": 2.9943083395199213e-06,
|
| 1380 |
+
"loss": 0.1692,
|
| 1381 |
+
"step": 19600
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.925018559762435,
|
| 1385 |
+
"grad_norm": 0.328125,
|
| 1386 |
+
"learning_rate": 2.4993813412521652e-06,
|
| 1387 |
+
"loss": 0.1708,
|
| 1388 |
+
"step": 19700
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.939866369710468,
|
| 1392 |
+
"grad_norm": 0.265625,
|
| 1393 |
+
"learning_rate": 2.00445434298441e-06,
|
| 1394 |
+
"loss": 0.1689,
|
| 1395 |
+
"step": 19800
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 2.9547141796585006,
|
| 1399 |
+
"grad_norm": 0.25,
|
| 1400 |
+
"learning_rate": 1.5095273447166542e-06,
|
| 1401 |
+
"loss": 0.1622,
|
| 1402 |
+
"step": 19900
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.9695619896065333,
|
| 1406 |
+
"grad_norm": 0.298828125,
|
| 1407 |
+
"learning_rate": 1.0146003464488989e-06,
|
| 1408 |
+
"loss": 0.1649,
|
| 1409 |
+
"step": 20000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 2.984409799554566,
|
| 1413 |
+
"grad_norm": 0.291015625,
|
| 1414 |
+
"learning_rate": 5.196733481811434e-07,
|
| 1415 |
+
"loss": 0.163,
|
| 1416 |
+
"step": 20100
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 2.9992576095025982,
|
| 1420 |
+
"grad_norm": 0.359375,
|
| 1421 |
+
"learning_rate": 2.4746349913387775e-08,
|
| 1422 |
+
"loss": 0.1617,
|
| 1423 |
+
"step": 20200
|
| 1424 |
+
}
|
| 1425 |
+
],
|
| 1426 |
+
"logging_steps": 100,
|
| 1427 |
+
"max_steps": 20205,
|
| 1428 |
+
"num_input_tokens_seen": 0,
|
| 1429 |
+
"num_train_epochs": 3,
|
| 1430 |
+
"save_steps": 1000,
|
| 1431 |
+
"stateful_callbacks": {
|
| 1432 |
+
"TrainerControl": {
|
| 1433 |
+
"args": {
|
| 1434 |
+
"should_epoch_stop": false,
|
| 1435 |
+
"should_evaluate": false,
|
| 1436 |
+
"should_log": false,
|
| 1437 |
+
"should_save": true,
|
| 1438 |
+
"should_training_stop": true
|
| 1439 |
+
},
|
| 1440 |
+
"attributes": {}
|
| 1441 |
+
}
|
| 1442 |
+
},
|
| 1443 |
+
"total_flos": 1.1772177365591194e+18,
|
| 1444 |
+
"train_batch_size": 1,
|
| 1445 |
+
"trial_name": null,
|
| 1446 |
+
"trial_params": null
|
| 1447 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88c498d78d918b36d7bfe364d212649599ca74142ed0b238e683d2dc18ea54fa
|
| 3 |
+
size 5280
|