metadata
license: bigcode-openrail-m
base_model: bigcode/starcoder
tags:
- generated_from_trainer
model-index:
- name: peft-lora-starcoder15B-v2-personal-copilot-A100-40GB-colab
results: []
library_name: peft
peft-lora-starcoder15B-v2-personal-copilot-A100-40GB-colab
This model is a fine-tuned version of bigcode/starcoder on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.3700
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
The following bitsandbytes quantization config was used during training:
- quant_method: bitsandbytes
- load_in_8bit: False
- load_in_4bit: True
- llm_int8_threshold: 6.0
- llm_int8_skip_modules: None
- llm_int8_enable_fp32_cpu_offload: False
- llm_int8_has_fp16_weight: False
- bnb_4bit_quant_type: nf4
- bnb_4bit_use_double_quant: True
- bnb_4bit_compute_dtype: bfloat16
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0005
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 30
- training_steps: 1000
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.6438 | 0.1 | 100 | 0.5596 |
| 0.5992 | 0.2 | 200 | 0.4886 |
| 0.6327 | 0.3 | 300 | 0.4319 |
| 0.5269 | 0.4 | 400 | 0.4077 |
| 0.4533 | 0.5 | 500 | 0.3990 |
| 0.4755 | 0.6 | 600 | 0.3852 |
| 0.4445 | 0.7 | 700 | 0.3735 |
| 0.5179 | 0.8 | 800 | 0.3721 |
| 0.4185 | 0.9 | 900 | 0.3706 |
| 0.4807 | 1.0 | 1000 | 0.3700 |
Framework versions
- PEFT 0.5.0.dev0
- Transformers 4.32.0.dev0
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3