Training in progress, step 5

README.md

@@ -1,17 +1,16 @@
 ---
-base_model: Qwen/Qwen3-0.6B
 library_name: transformers
-model_name: sft_full_0.6_pre
+model_name: dpo_full_1.7
 tags:
 - generated_from_trainer
 - trl
-- sft
+- dpo
 licence: license
 ---
 
-# Model Card for sft_full_0.6_pre
+# Model Card for dpo_full_1.7
 
-This model is a fine-tuned version of [Qwen/Qwen3-0.6B](https://huggingface.co/Qwen/Qwen3-0.6B).
+This model is a fine-tuned version of [None](https://huggingface.co/None).
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start
@@ -20,17 +19,17 @@ It has been trained using [TRL](https://github.com/huggingface/trl).
 from transformers import pipeline
 
 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="SaminSkyfall/sft_full_0.6_pre", device="cuda")
+generator = pipeline("text-generation", model="SaminSkyfall/dpo_full_1.7", device="cuda")
 output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
 print(output["generated_text"])
 ```
 
 ## Training procedure
 
-[<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/samin-skyfall-ai/huggingface/runs/3d7ylu2c) 
+[<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/samin-skyfall-ai/huggingface/runs/lc17gapu) 
 
 
-This model was trained with SFT.
+This model was trained with DPO, a method introduced in [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://huggingface.co/papers/2305.18290).
 
 ### Framework versions
 
@@ -42,7 +41,18 @@ This model was trained with SFT.
 
 ## Citations
 
+Cite DPO as:
 
+```bibtex
+@inproceedings{rafailov2023direct,
+    title        = {{Direct Preference Optimization: Your Language Model is Secretly a Reward Model}},
+    author       = {Rafael Rafailov and Archit Sharma and Eric Mitchell and Christopher D. Manning and Stefano Ermon and Chelsea Finn},
+    year         = 2023,
+    booktitle    = {Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems 2023, NeurIPS 2023, New Orleans, LA, USA, December 10 - 16, 2023},
+    url          = {http://papers.nips.cc/paper_files/paper/2023/hash/a85b405ed65c6477a4fe8302b5e06ce7-Abstract-Conference.html},
+    editor       = {Alice Oh and Tristan Naumann and Amir Globerson and Kate Saenko and Moritz Hardt and Sergey Levine},
+}
+```
 
 Cite TRL as:
     

adapter_config.json

@@ -1,7 +1,7 @@
 {
   "alpha_pattern": {},
   "auto_mapping": null,
-  "base_model_name_or_path": "Qwen/Qwen3-1.7B",
+  "base_model_name_or_path": "Qwen/Qwen3-0.6B",
   "bias": "none",
   "corda_config": null,
   "eva_config": null,
@@ -18,21 +18,24 @@
   "lora_dropout": 0.0,
   "megatron_config": null,
   "megatron_core": "megatron.core",
-  "modules_to_save": null,
+  "modules_to_save": [
+    "classifier",
+    "score"
+  ],
   "peft_type": "LORA",
   "r": 32,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
     "up_proj",
-    "q_proj",
     "gate_proj",
     "o_proj",
     "k_proj",
     "down_proj",
-    "v_proj"
+    "v_proj",
+    "q_proj"
   ],
-  "task_type": "CAUSAL_LM",
+  "task_type": "SEQ_CLS",
   "trainable_token_indices": null,
   "use_dora": false,
   "use_rslora": false

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fc980bc3a09519384af5e0b77d15073ef269a7af6e5192ff97486fa1d14d071c
-size 139512976
+oid sha256:7b4852f593289d8d754fb266ca714000259357a59ed79306d2de4f8dba95c8d0
+size 80796648

config.json

@@ -23,7 +23,7 @@
   "rope_theta": 1000000,
   "sliding_window": null,
   "tie_word_embeddings": true,
-  "torch_dtype": "float16",
+  "torch_dtype": "float32",
   "transformers_version": "4.51.3",
   "use_cache": true,
   "use_sliding_window": false,

model-00001-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:174c6d288e87934a230721c5d098f7e5437de1c52f163a0b628628b9d70329d5
+oid sha256:22ec6d375dfdbdaa1a7fa473ec529a3f4bb6b5f096aa185eb87bec70ef2dc35b
 size 4969539560

model-00002-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aa6d7b6fa4e7b23181588fa473b007f5160094762338ba38b346688fefd8d1f0
+oid sha256:753d24c6c821e2a4ddd49039983732d238c3f062c6cd92361141e9f9d4e843fa
 size 1912795688

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:888bfd9d9a3a3086b0ec0f1f1c63d1c7f032ef68e2a3082a9d85326812373d0d
+oid sha256:ce90abc86c871f836a6032a8a14fa996159108621f757ccb76c6a4b82a4a59fd
 size 6033