End of training

README.md

@@ -1,14 +1,14 @@
 ---
 library_name: peft
 license: apache-2.0
-base_model: unsloth/Qwen2-0.5B-Instruct
+base_model: unsloth/codegemma-2b
 tags:
 - axolotl
 - generated_from_trainer
 datasets:
-- Aivesa/dataset_63314436-7a2d-452f-be27-72d9982651a1
+- Aivesa/dataset_f7765789-78a9-498c-a963-5178a84d33b4
 model-index:
-- name: 53638570-6e01-4257-8d08-795d436a977c
+- name: 349aacef-3660-4206-b561-6d7b5d5a9b62
   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: unsloth/Qwen2-0.5B-Instruct
+base_model: unsloth/codegemma-2b
 bf16: auto
 chat_template: llama3
 dataset_prepared_path: /workspace/axolotl/data/prepared
 datasets:
 - ds_type: json
   format: custom
-  path: Aivesa/dataset_63314436-7a2d-452f-be27-72d9982651a1
+  path: Aivesa/dataset_f7765789-78a9-498c-a963-5178a84d33b4
   type:
-    field_input: background
-    field_instruction: rationale
-    field_output: evaluation
+    field_input: hypothesis
+    field_instruction: prompt
+    field_output: transcription
     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/53638570-6e01-4257-8d08-795d436a977c
+hub_model_id: Aivesa/349aacef-3660-4206-b561-6d7b5d5a9b62
 hub_private_repo: true
 hub_repo: null
 hub_strategy: checkpoint
@@ -88,10 +88,10 @@ use_accelerate: true
 val_set_size: 0.05
 wandb_entity: null
 wandb_mode: online
-wandb_name: 63314436-7a2d-452f-be27-72d9982651a1
+wandb_name: f7765789-78a9-498c-a963-5178a84d33b4
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
-wandb_runid: 63314436-7a2d-452f-be27-72d9982651a1
+wandb_runid: f7765789-78a9-498c-a963-5178a84d33b4
 warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
@@ -100,11 +100,11 @@ xformers_attention: null
 
 </details><br>
 
-# 53638570-6e01-4257-8d08-795d436a977c
+# 349aacef-3660-4206-b561-6d7b5d5a9b62
 
-This model is a fine-tuned version of [unsloth/Qwen2-0.5B-Instruct](https://huggingface.co/unsloth/Qwen2-0.5B-Instruct) on the Aivesa/dataset_63314436-7a2d-452f-be27-72d9982651a1 dataset.
+This model is a fine-tuned version of [unsloth/codegemma-2b](https://huggingface.co/unsloth/codegemma-2b) on the Aivesa/dataset_f7765789-78a9-498c-a963-5178a84d33b4 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.7616
+- Loss: 0.7542
 
 ## Model description
 
@@ -138,9 +138,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 2.7241        | 0.0013 | 3    | 2.4370          |
-| 2.324         | 0.0025 | 6    | 2.2187          |
-| 1.9242        | 0.0038 | 9    | 1.7616          |
+| 1.9715        | 0.0001 | 3    | 2.2125          |
+| 1.4737        | 0.0002 | 6    | 1.6327          |
+| 1.8642        | 0.0002 | 9    | 0.7542          |
 
 
 ### Framework versions