End of training

README.md

@@ -2,9 +2,15 @@
 license: apache-2.0
 base_model: martimfasantos/tinyllama-1.1b-sum-sft-full_old
 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_LR2e-8_2epochs_old
   results: []
@@ -15,7 +21,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # tinyllama-1.1b-sum-dpo-full_LR2e-8_2epochs_old
 
-This model is a fine-tuned version of [martimfasantos/tinyllama-1.1b-sum-sft-full_old](https://huggingface.co/martimfasantos/tinyllama-1.1b-sum-sft-full_old) on an unknown dataset.
+This model is a fine-tuned version of [martimfasantos/tinyllama-1.1b-sum-sft-full_old](https://huggingface.co/martimfasantos/tinyllama-1.1b-sum-sft-full_old) on the openai/summarize_from_feedback dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.6901
 - Rewards/chosen: -0.0088

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 2.0,
+    "eval_logits/chosen": -3.11618709564209,
+    "eval_logits/rejected": -3.1104962825775146,
+    "eval_logps/chosen": -59.58960723876953,
+    "eval_logps/rejected": -64.70088958740234,
+    "eval_loss": 0.6901275515556335,
+    "eval_rewards/accuracies": 0.589219331741333,
+    "eval_rewards/chosen": -0.00877712108194828,
+    "eval_rewards/margins": 0.006430591456592083,
+    "eval_rewards/rejected": -0.015207710675895214,
+    "eval_runtime": 383.5579,
+    "eval_samples": 4304,
+    "eval_samples_per_second": 11.221,
+    "eval_steps_per_second": 1.403,
     "total_flos": 0.0,
     "train_loss": 0.6883086910911629,
     "train_runtime": 95005.2512,

config.json

@@ -24,6 +24,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "float32",
   "transformers_version": "4.41.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 2.0,
+    "eval_logits/chosen": -3.11618709564209,
+    "eval_logits/rejected": -3.1104962825775146,
+    "eval_logps/chosen": -59.58960723876953,
+    "eval_logps/rejected": -64.70088958740234,
+    "eval_loss": 0.6901275515556335,
+    "eval_rewards/accuracies": 0.589219331741333,
+    "eval_rewards/chosen": -0.00877712108194828,
+    "eval_rewards/margins": 0.006430591456592083,
+    "eval_rewards/rejected": -0.015207710675895214,
+    "eval_runtime": 383.5579,
+    "eval_samples": 4304,
+    "eval_samples_per_second": 11.221,
+    "eval_steps_per_second": 1.403
+}

runs/Jun21_02-49-17_poseidon/events.out.tfevents.1719034213.poseidon.136652.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8396d50e1e23f9ee651d16bc5ef4609d5a23d4d0c0c7c689d8fa7974cd3e2c22
+size 828