End of training

README.md

@@ -2,9 +2,15 @@
 license: apache-2.0
 base_model: martimfasantos/tinyllama-1.1b-sum-sft-full
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- openai/summarize_from_feedback
 model-index:
 - name: tinyllama-1.1b-sum-dpo-full
   results: []
@@ -15,7 +21,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # tinyllama-1.1b-sum-dpo-full
 
-This model is a fine-tuned version of [martimfasantos/tinyllama-1.1b-sum-sft-full](https://huggingface.co/martimfasantos/tinyllama-1.1b-sum-sft-full) on an unknown dataset.
+This model is a fine-tuned version of [martimfasantos/tinyllama-1.1b-sum-sft-full](https://huggingface.co/martimfasantos/tinyllama-1.1b-sum-sft-full) on the openai/summarize_from_feedback dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.6342
 - Rewards/chosen: -1.8568

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 2.0,
+    "eval_logits/chosen": -1.4189956188201904,
+    "eval_logits/rejected": -1.3920025825500488,
+    "eval_logps/chosen": -244.38746643066406,
+    "eval_logps/rejected": -295.19293212890625,
+    "eval_loss": 0.6342188119888306,
+    "eval_rewards/accuracies": 0.6579925417900085,
+    "eval_rewards/chosen": -1.8568360805511475,
+    "eval_rewards/margins": 0.4635196328163147,
+    "eval_rewards/rejected": -2.3203558921813965,
+    "eval_runtime": 354.8785,
+    "eval_samples": 4304,
+    "eval_samples_per_second": 12.128,
+    "eval_steps_per_second": 1.516,
     "train_loss": 0.5042109644922366,
     "train_runtime": 89019.0317,
     "train_samples": 92858,

config.json

@@ -23,6 +23,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.39.3",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 2.0,
+    "eval_logits/chosen": -1.4189956188201904,
+    "eval_logits/rejected": -1.3920025825500488,
+    "eval_logps/chosen": -244.38746643066406,
+    "eval_logps/rejected": -295.19293212890625,
+    "eval_loss": 0.6342188119888306,
+    "eval_rewards/accuracies": 0.6579925417900085,
+    "eval_rewards/chosen": -1.8568360805511475,
+    "eval_rewards/margins": 0.4635196328163147,
+    "eval_rewards/rejected": -2.3203558921813965,
+    "eval_runtime": 354.8785,
+    "eval_samples": 4304,
+    "eval_samples_per_second": 12.128,
+    "eval_steps_per_second": 1.516
+}

runs/May15_16-59-39_poseidon/events.out.tfevents.1715882392.poseidon.3676190.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5f1bd12d17c247b9b8f9dc37b41a4ce5aff2e8543e15b198adbc9db02f6cab9
+size 828