End of training

README.md

@@ -2,9 +2,15 @@
 license: gemma
 base_model: ale-bay/zephyr-2b-gemma-sft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- argilla/dpo-mix-7k
 model-index:
 - name: zephyr-2b-gemma-sft-dpo
   results: []
@@ -16,17 +22,17 @@ should probably proofread and complete it, then remove this comment. -->
 [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://zebra.wandb.io/cto/distillm/runs/shrbgd04)
 # zephyr-2b-gemma-sft-dpo
 
-This model is a fine-tuned version of [ale-bay/zephyr-2b-gemma-sft](https://huggingface.co/ale-bay/zephyr-2b-gemma-sft) on an unknown dataset.
+This model is a fine-tuned version of [ale-bay/zephyr-2b-gemma-sft](https://huggingface.co/ale-bay/zephyr-2b-gemma-sft) on the argilla/dpo-mix-7k dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5960
-- Rewards/chosen: 0.2261
-- Rewards/rejected: -0.0336
+- Loss: 0.5950
+- Rewards/chosen: 0.2219
+- Rewards/rejected: -0.0437
 - Rewards/accuracies: 0.7188
-- Rewards/margins: 0.2597
-- Logps/rejected: -363.7130
-- Logps/chosen: -363.7419
+- Rewards/margins: 0.2656
+- Logps/rejected: -363.9149
+- Logps/chosen: -363.8262
 - Logits/rejected: -15.9487
-- Logits/chosen: -16.5860
+- Logits/chosen: -16.5867
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.971563981042654,
+    "eval_logits/chosen": -16.5866641998291,
+    "eval_logits/rejected": -15.948689460754395,
+    "eval_logps/chosen": -363.826171875,
+    "eval_logps/rejected": -363.9148864746094,
+    "eval_loss": 0.5950456857681274,
+    "eval_rewards/accuracies": 0.71875,
+    "eval_rewards/chosen": 0.22188298404216766,
+    "eval_rewards/margins": 0.26559606194496155,
+    "eval_rewards/rejected": -0.043713077902793884,
+    "eval_runtime": 6.8121,
+    "eval_samples": 750,
+    "eval_samples_per_second": 110.098,
+    "eval_steps_per_second": 3.523,
     "total_flos": 0.0,
     "train_loss": 0.6262502945386447,
     "train_runtime": 431.1142,

config.json

@@ -24,6 +24,6 @@
   "rope_theta": 10000.0,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.42.3",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 256000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 1.971563981042654,
+    "eval_logits/chosen": -16.5866641998291,
+    "eval_logits/rejected": -15.948689460754395,
+    "eval_logps/chosen": -363.826171875,
+    "eval_logps/rejected": -363.9148864746094,
+    "eval_loss": 0.5950456857681274,
+    "eval_rewards/accuracies": 0.71875,
+    "eval_rewards/chosen": 0.22188298404216766,
+    "eval_rewards/margins": 0.26559606194496155,
+    "eval_rewards/rejected": -0.043713077902793884,
+    "eval_runtime": 6.8121,
+    "eval_samples": 750,
+    "eval_samples_per_second": 110.098,
+    "eval_steps_per_second": 3.523
+}

runs/Jul11_11-23-58_ale-distillm-8-0-0/events.out.tfevents.1720694314.ale-distillm-8-0-0.3691.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b741a841a029947996bc919ab4020585f665fd2cb0f8c5d14d6cd5bd9c9174d2
+size 815