End of training

README.md

@@ -1,14 +1,14 @@
 ---
 library_name: peft
-license: gemma
-base_model: zake7749/gemma-2-2b-it-chinese-kyara-dpo
+license: apache-2.0
+base_model: EleutherAI/pythia-410m-deduped
 tags:
 - axolotl
 - generated_from_trainer
 datasets:
-- Paladiso/dataset_48ec0ffb-9278-423b-bb92-4d2d7ba40108
+- Paladiso/dataset_f5aaf6b6-8b8a-4d62-a003-0a4288146cc5
 model-index:
-- name: 8ccdbc24-1387-4605-99fc-ca461a35399b
+- name: b0ed7ce7-ba23-491f-ab37-49149ecf7349
   results: []
 ---
 
@@ -21,18 +21,18 @@ should probably proofread and complete it, then remove this comment. -->
 axolotl version: `0.6.0`
 ```yaml
 adapter: lora
-base_model: zake7749/gemma-2-2b-it-chinese-kyara-dpo
+base_model: EleutherAI/pythia-410m-deduped
 bf16: auto
 chat_template: llama3
 dataset_prepared_path: /workspace/axolotl/data/prepared
 datasets:
 - ds_type: json
   format: custom
-  path: Paladiso/dataset_48ec0ffb-9278-423b-bb92-4d2d7ba40108
+  path: Paladiso/dataset_f5aaf6b6-8b8a-4d62-a003-0a4288146cc5
   type:
-    field_input: authors
-    field_instruction: type
-    field_output: title
+    field_input: title
+    field_instruction: query
+    field_output: positive
     system_format: '{system}'
     system_prompt: ''
 debug: null
@@ -48,7 +48,7 @@ fsdp_config: null
 gradient_accumulation_steps: 4
 gradient_checkpointing: false
 group_by_length: false
-hub_model_id: Paladiso/8ccdbc24-1387-4605-99fc-ca461a35399b
+hub_model_id: Paladiso/b0ed7ce7-ba23-491f-ab37-49149ecf7349
 hub_private_repo: true
 hub_repo: null
 hub_strategy: checkpoint
@@ -79,6 +79,8 @@ sample_packing: false
 save_safetensors: true
 saves_per_epoch: 4
 sequence_len: 512
+special_tokens:
+  pad_token: <|endoftext|>
 strict: false
 tf32: false
 tokenizer_type: AutoTokenizer
@@ -88,10 +90,10 @@ use_accelerate: true
 val_set_size: 0.05
 wandb_entity: null
 wandb_mode: online
-wandb_name: 48ec0ffb-9278-423b-bb92-4d2d7ba40108
+wandb_name: f5aaf6b6-8b8a-4d62-a003-0a4288146cc5
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
-wandb_runid: 48ec0ffb-9278-423b-bb92-4d2d7ba40108
+wandb_runid: f5aaf6b6-8b8a-4d62-a003-0a4288146cc5
 warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
@@ -100,11 +102,11 @@ xformers_attention: null
 
 </details><br>
 
-# 8ccdbc24-1387-4605-99fc-ca461a35399b
+# b0ed7ce7-ba23-491f-ab37-49149ecf7349
 
-This model is a fine-tuned version of [zake7749/gemma-2-2b-it-chinese-kyara-dpo](https://huggingface.co/zake7749/gemma-2-2b-it-chinese-kyara-dpo) on the Paladiso/dataset_48ec0ffb-9278-423b-bb92-4d2d7ba40108 dataset.
+This model is a fine-tuned version of [EleutherAI/pythia-410m-deduped](https://huggingface.co/EleutherAI/pythia-410m-deduped) on the Paladiso/dataset_f5aaf6b6-8b8a-4d62-a003-0a4288146cc5 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 4.6824
+- Loss: 3.4121
 
 ## Model description
 
@@ -138,9 +140,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 7.3026        | 0.0007 | 3    | 7.5792          |
-| 5.8698        | 0.0014 | 6    | 5.9740          |
-| 5.1532        | 0.0020 | 9    | 4.6824          |
+| 13.5435       | 0.0003 | 3    | 3.4388          |
+| 13.2872       | 0.0006 | 6    | 3.4335          |
+| 13.6718       | 0.0008 | 9    | 3.4121          |
 
 
 ### Framework versions