HeZ/phi-3-mini-4k-ft-code-gen-v2

README.md

@@ -1,67 +1,57 @@
 ---
 base_model: microsoft/Phi-3-mini-4k-instruct
-library_name: peft
-license: mit
+library_name: transformers
+model_name: phi-3-mini-LoRA
 tags:
+- generated_from_trainer
 - trl
 - sft
-- generated_from_trainer
-model-index:
-- name: phi-3-mini-LoRA
-  results: []
+licence: license
 ---
 
-<!-- This model card has been generated automatically according to the information the Trainer had access to. You
-should probably proofread and complete it, then remove this comment. -->
-
-# phi-3-mini-LoRA
-
-This model is a fine-tuned version of [microsoft/Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct) on an unknown dataset.
-It achieves the following results on the evaluation set:
-- Loss: 0.5759
-
-## Model description
+# Model Card for phi-3-mini-LoRA
 
-More information needed
+This model is a fine-tuned version of [microsoft/Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct).
+It has been trained using [TRL](https://github.com/huggingface/trl).
 
-## Intended uses & limitations
+## Quick start
 
-More information needed
+```python
+from transformers import pipeline
 
-## Training and evaluation data
-
-More information needed
+question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
+generator = pipeline("text-generation", model="HeZ/phi-3-mini-LoRA", device="cuda")
+output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
+print(output["generated_text"])
+```
 
 ## Training procedure
 
-### Training hyperparameters
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/hezhang33/Phi3-mini-ft-python-code/runs/sggnnts3)
+
+This model was trained with SFT.
 
-The following hyperparameters were used during training:
-- learning_rate: 0.0001
-- train_batch_size: 1
-- eval_batch_size: 1
-- seed: 42
-- gradient_accumulation_steps: 40
-- total_train_batch_size: 40
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: linear
-- lr_scheduler_warmup_ratio: 0.1
-- num_epochs: 1
+### Framework versions
 
-### Training results
+- TRL: 0.12.2
+- Transformers: 4.46.3
+- Pytorch: 2.7.1
+- Datasets: 4.0.0
+- Tokenizers: 0.20.3
 
-| Training Loss | Epoch  | Step | Validation Loss |
-|:-------------:|:------:|:----:|:---------------:|
-| 0.7817        | 0.2262 | 100  | 0.5929          |
-| 0.5969        | 0.4525 | 200  | 0.5816          |
-| 0.5848        | 0.6787 | 300  | 0.5775          |
-| 0.581         | 0.9049 | 400  | 0.5759          |
+## Citations
 
 
-### Framework versions
 
-- PEFT 0.13.0
-- Transformers 4.45.1
-- Pytorch 2.4.1+cu121
-- Datasets 3.0.1
-- Tokenizers 0.20.0
+Cite TRL as:
+    
+```bibtex
+@misc{vonwerra2022trl,
+	title        = {{TRL: Transformer Reinforcement Learning}},
+	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
+	year         = 2020,
+	journal      = {GitHub repository},
+	publisher    = {GitHub},
+	howpublished = {\url{https://github.com/huggingface/trl}}
+}
+```

adapter_config.json

@@ -3,6 +3,9 @@
   "auto_mapping": null,
   "base_model_name_or_path": "microsoft/Phi-3-mini-4k-instruct",
   "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": null,
   "fan_in_fan_out": false,
   "inference_mode": true,
   "init_lora_weights": true,
@@ -11,24 +14,28 @@
   "layers_to_transform": null,
   "loftq_config": {},
   "lora_alpha": 16,
+  "lora_bias": false,
   "lora_dropout": 0.05,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
+  "qalora_group_size": 16,
   "r": 16,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
+    "down_proj",
+    "o_proj",
+    "up_proj",
     "v_proj",
     "gate_proj",
-    "o_proj",
     "k_proj",
-    "up_proj",
-    "down_proj"
+    "q_proj"
   ],
   "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
   "use_dora": false,
+  "use_qalora": false,
   "use_rslora": false
 }

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eb5611f02feac3c0646a003d18c905c2b55c2fb97809601200ce45b80ebb2fd0
+oid sha256:dd3f25cd467e644f3e8cd13b8f3bbd0aea947bec687a881549b03112b8863f21
 size 35668592

tokenizer.json

@@ -1,6 +1,11 @@
 {
   "version": "1.0",
-  "truncation": null,
+  "truncation": {
+    "direction": "Right",
+    "max_length": 512,
+    "strategy": "LongestFirst",
+    "stride": 0
+  },
   "padding": null,
   "added_tokens": [
     {

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a1c27ad626e45d66f176593b06c39cbc1489497460e6b488a616c15f7e7ba720
-size 5432
+oid sha256:54eb70c6f1ba331a040983c06f83efb1f0c65788b5e67e63fba82ca37d395d7b
+size 5969