aidev08/+phi-3-mini-4k-instruct-qlora-MathInstruct+adapter

README.md

@@ -1,73 +1,57 @@
 ---
 base_model: microsoft/Phi-3-mini-4k-instruct
-library_name: peft
-license: mit
+library_name: transformers
+model_name: phi-3-mini-QLoRA
 tags:
+- generated_from_trainer
 - trl
 - sft
-- generated_from_trainer
-model-index:
-- name: phi-3-mini-QLoRA
-  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-QLoRA
-
-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: 2.0585
+# Model Card for phi-3-mini-QLoRA
 
-## Model description
+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).
 
-More information needed
+## Quick start
 
-## Intended uses & limitations
+```python
+from transformers import pipeline
 
-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="aidev08/phi-3-mini-QLoRA", device="cuda")
+output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
+print(output["generated_text"])
+```
 
-## Training and evaluation data
+## Training procedure
 
-More information needed
 
-## Training procedure
 
-### Training hyperparameters
+This model was trained with SFT.
 
-The following hyperparameters were used during training:
-- learning_rate: 0.0001
-- train_batch_size: 8
-- eval_batch_size: 8
-- seed: 42
-- gradient_accumulation_steps: 4
-- total_train_batch_size: 32
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: linear
-- lr_scheduler_warmup_ratio: 0.1
-- num_epochs: 3
+### Framework versions
 
-### Training results
+- TRL: 0.12.0
+- Transformers: 4.46.1
+- Pytorch: 2.5.1
+- Datasets: 3.1.0
+- Tokenizers: 0.20.1
 
-| Training Loss | Epoch  | Step | Validation Loss |
-|:-------------:|:------:|:----:|:---------------:|
-| 2.4323        | 0.2878 | 30   | 2.2759          |
-| 2.206         | 0.5755 | 60   | 2.1256          |
-| 2.1409        | 0.8633 | 90   | 2.1004          |
-| 2.0985        | 1.1511 | 120  | 2.0887          |
-| 2.0991        | 1.4388 | 150  | 2.0801          |
-| 2.1           | 1.7266 | 180  | 2.0723          |
-| 2.095         | 2.0144 | 210  | 2.0675          |
-| 2.073         | 2.3022 | 240  | 2.0628          |
-| 2.0729        | 2.5899 | 270  | 2.0605          |
-| 2.0864        | 2.8777 | 300  | 2.0585          |
+## Citations
 
 
-### Framework versions
 
-- PEFT 0.12.0
-- Transformers 4.44.2
-- Pytorch 2.4.1+cu121
-- Datasets 2.21.0
-- Tokenizers 0.19.1
+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

@@ -20,13 +20,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "o_proj",
-    "gate_proj",
-    "k_proj",
-    "up_proj",
     "v_proj",
+    "down_proj",
+    "k_proj",
     "q_proj",
-    "down_proj"
+    "gate_proj",
+    "o_proj",
+    "up_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eebcb95f27c41c83978b244295e56b7cb1ff3b6966b68fcee6f26107d6aa2da9
+oid sha256:5b9eb01ff27356e7a496fb1a4bdbdf901f890b48a626b6400ea1463b5e045fb1
 size 35668592

runs/Nov02_14-59-03_7b48f43707b1/events.out.tfevents.1730559655.7b48f43707b1.5437.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:485156c8d6a11a0442002618e2e87013551ee488cd6863bb3b205dd301d6ee7c
+size 6197

runs/Nov02_15-02-49_7b48f43707b1/events.out.tfevents.1730559776.7b48f43707b1.5437.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5ff33a875a85944a0d49c66df89ad58b28077972c71cd9da8209d25d9b48a1f
+size 28802

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9a220015c2b7635f22328bb29952e8da83b5af58206d6f2fd7d8a2e16553eb09
-size 5432
+oid sha256:b985f3da691b40dca1cdcf4928e2b6dde5a2e56f2e864c1fd4413748aa038543
+size 5496