Model save

README.md

@@ -1,7 +1,7 @@
 ---
-base_model: RyanYr/reflect_mini8Bit_om2-460k_sft-t1
+base_model: RyanYr/reflect_ministral8Bit_om2_sft-t2_lr.5-6
 library_name: transformers
-model_name: reflect_mini8B_Om2SftT1-Om2G8kOm2AgIpsdpIter1T1_b0.5
+model_name: reflect_mini8B_Om2SftT2_Om2G8kOm2AgIpsdpIter1T02_b1.0
 tags:
 - generated_from_trainer
 - trl
@@ -9,9 +9,9 @@ tags:
 licence: license
 ---
 
-# Model Card for reflect_mini8B_Om2SftT1-Om2G8kOm2AgIpsdpIter1T1_b0.5
+# Model Card for reflect_mini8B_Om2SftT2_Om2G8kOm2AgIpsdpIter1T02_b1.0
 
-This model is a fine-tuned version of [RyanYr/reflect_mini8Bit_om2-460k_sft-t1](https://huggingface.co/RyanYr/reflect_mini8Bit_om2-460k_sft-t1).
+This model is a fine-tuned version of [RyanYr/reflect_ministral8Bit_om2_sft-t2_lr.5-6](https://huggingface.co/RyanYr/reflect_ministral8Bit_om2_sft-t2_lr.5-6).
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start
@@ -20,14 +20,14 @@ 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="RyanYr/reflect_mini8B_Om2SftT1-Om2G8kOm2AgIpsdpIter1T1_b0.5", device="cuda")
+generator = pipeline("text-generation", model="RyanYr/reflect_mini8B_Om2SftT2_Om2G8kOm2AgIpsdpIter1T02_b1.0", 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/yyr/huggingface/runs/0jnfhtyj)
+[<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/yyr/huggingface/runs/8zzk5a6s)
 
 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).
 

last_checkpoint/config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "RyanYr/reflect_mini8Bit_om2-460k_sft-t1",
+  "_name_or_path": "RyanYr/reflect_ministral8Bit_om2_sft-t2_lr.5-6",
   "architectures": [
     "MistralForCausalLM"
   ],

last_checkpoint/model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:525ff610191554c89a15b5ef18a2fc3196519f5878fb5978b1499328c7fafaa0
+oid sha256:c6bdd8545c9ad01574273dd0b3c8deeef5ee81d1b6ac3edd60be5072effa3632
 size 4983016096

last_checkpoint/model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a05f6450585d459c438497c9a94318a70c470d81c786163c5c9f9e78f8a214a3
+oid sha256:cc85fb01321d7b15b030ffef518f86f6736d11b9248e55220d863a0c51227686
 size 4999836776

last_checkpoint/model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:651fa244689ad64bf4a6463df0c9cebc3b5aa4a6c43a495e8f8d82c4285198e4
+oid sha256:39e68854d7294897189b9fc00c8e53ab23ac18af7778ec9a64c6a13832d95d04
 size 4983067960

last_checkpoint/model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a8b2d574e2f43b04f4f7276e31c1967110c642020c8774dbb117ae7341087e12
+oid sha256:cd398d5a65a825334f17678e08d8313e3c7975c5523e3d081787973a7a049db9
 size 1073750144

last_checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f8e59085d8d6d8a265bdac9bbd1110865c36847f1d93dc1e4b5be688a06556a8
+oid sha256:1d4fa864da1ac1fe7f4bff3f2d1c0068a8366f7c27cefb89717f8fefeb9fa2f5
 size 8056