End of training

README.md

@@ -1,14 +1,14 @@
 ---
 library_name: peft
-license: apache-2.0
-base_model: JackFram/llama-68m
+license: other
+base_model: facebook/opt-1.3b
 tags:
 - axolotl
 - generated_from_trainer
 datasets:
-- Aivesa/dataset_018784f0-a53d-4511-af7a-e966236fc582
+- Aivesa/dataset_46dd44b3-e4e4-4c44-8605-e8e8b6dd956e
 model-index:
-- name: 1ecfbdde-71a6-46a4-8ea6-5fb1b3315696
+- name: a05b476c-463b-4946-89c4-d93aca81d7e7
   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: JackFram/llama-68m
+base_model: facebook/opt-1.3b
 bf16: auto
 chat_template: llama3
 dataset_prepared_path: /workspace/axolotl/data/prepared
 datasets:
 - ds_type: json
   format: custom
-  path: Aivesa/dataset_018784f0-a53d-4511-af7a-e966236fc582
+  path: Aivesa/dataset_46dd44b3-e4e4-4c44-8605-e8e8b6dd956e
   type:
     field_input: context
-    field_instruction: source
-    field_output: reference_original
+    field_instruction: question
+    field_output: answers
     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: Aivesa/1ecfbdde-71a6-46a4-8ea6-5fb1b3315696
+hub_model_id: Aivesa/a05b476c-463b-4946-89c4-d93aca81d7e7
 hub_private_repo: true
 hub_repo: null
 hub_strategy: checkpoint
@@ -79,8 +79,6 @@ sample_packing: false
 save_safetensors: true
 saves_per_epoch: 4
 sequence_len: 512
-special_tokens:
-  pad_token: </s>
 strict: false
 tf32: false
 tokenizer_type: AutoTokenizer
@@ -90,10 +88,10 @@ use_accelerate: true
 val_set_size: 0.05
 wandb_entity: null
 wandb_mode: online
-wandb_name: 018784f0-a53d-4511-af7a-e966236fc582
+wandb_name: 46dd44b3-e4e4-4c44-8605-e8e8b6dd956e
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
-wandb_runid: 018784f0-a53d-4511-af7a-e966236fc582
+wandb_runid: 46dd44b3-e4e4-4c44-8605-e8e8b6dd956e
 warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
@@ -102,11 +100,11 @@ xformers_attention: null
 
 </details><br>
 
-# 1ecfbdde-71a6-46a4-8ea6-5fb1b3315696
+# a05b476c-463b-4946-89c4-d93aca81d7e7
 
-This model is a fine-tuned version of [JackFram/llama-68m](https://huggingface.co/JackFram/llama-68m) on the Aivesa/dataset_018784f0-a53d-4511-af7a-e966236fc582 dataset.
+This model is a fine-tuned version of [facebook/opt-1.3b](https://huggingface.co/facebook/opt-1.3b) on the Aivesa/dataset_46dd44b3-e4e4-4c44-8605-e8e8b6dd956e dataset.
 It achieves the following results on the evaluation set:
-- Loss: 3.0933
+- Loss: 2.6826
 
 ## Model description
 
@@ -140,9 +138,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 3.1687        | 0.0149 | 3    | 3.1479          |
-| 2.8454        | 0.0299 | 6    | 3.1350          |
-| 3.1911        | 0.0448 | 9    | 3.0933          |
+| 15.8191       | 0.0003 | 3    | 4.1197          |
+| 17.6309       | 0.0005 | 6    | 3.6521          |
+| 11.7024       | 0.0008 | 9    | 2.6826          |
 
 
 ### Framework versions