End of training

README.md

@@ -1,13 +1,14 @@
 ---
 library_name: peft
-base_model: katuni4ka/tiny-random-dbrx
+license: apache-2.0
+base_model: openlm-research/open_llama_3b
 tags:
 - axolotl
 - generated_from_trainer
 datasets:
-- Paladiso/dataset_91d2e9f8-99da-4b30-8acb-ee8b1520d1bc
+- Paladiso/dataset_b98c9df0-5280-40e9-9125-37aafc6279e9
 model-index:
-- name: b6b31df9-7c59-4ee2-8a19-e3cd1fd58249
+- name: 369db93c-b0b2-4862-895e-e55b2f44da91
   results: []
 ---
 
@@ -20,18 +21,18 @@ should probably proofread and complete it, then remove this comment. -->
 axolotl version: `0.6.0`
 ```yaml
 adapter: lora
-base_model: katuni4ka/tiny-random-dbrx
+base_model: openlm-research/open_llama_3b
 bf16: auto
 chat_template: llama3
 dataset_prepared_path: /workspace/axolotl/data/prepared
 datasets:
 - ds_type: json
   format: custom
-  path: Paladiso/dataset_91d2e9f8-99da-4b30-8acb-ee8b1520d1bc
+  path: Paladiso/dataset_b98c9df0-5280-40e9-9125-37aafc6279e9
   type:
-    field_input: context
-    field_instruction: instruction
-    field_output: response
+    field_input: sentence1
+    field_instruction: genre
+    field_output: sentence2
     system_format: '{system}'
     system_prompt: ''
 debug: null
@@ -47,7 +48,7 @@ fsdp_config: null
 gradient_accumulation_steps: 4
 gradient_checkpointing: false
 group_by_length: false
-hub_model_id: Paladiso/b6b31df9-7c59-4ee2-8a19-e3cd1fd58249
+hub_model_id: Paladiso/369db93c-b0b2-4862-895e-e55b2f44da91
 hub_private_repo: true
 hub_repo: null
 hub_strategy: checkpoint
@@ -78,6 +79,8 @@ 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
@@ -87,10 +90,10 @@ use_accelerate: true
 val_set_size: 0.05
 wandb_entity: null
 wandb_mode: online
-wandb_name: 91d2e9f8-99da-4b30-8acb-ee8b1520d1bc
+wandb_name: b98c9df0-5280-40e9-9125-37aafc6279e9
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
-wandb_runid: 91d2e9f8-99da-4b30-8acb-ee8b1520d1bc
+wandb_runid: b98c9df0-5280-40e9-9125-37aafc6279e9
 warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
@@ -99,11 +102,11 @@ xformers_attention: null
 
 </details><br>
 
-# b6b31df9-7c59-4ee2-8a19-e3cd1fd58249
+# 369db93c-b0b2-4862-895e-e55b2f44da91
 
-This model is a fine-tuned version of [katuni4ka/tiny-random-dbrx](https://huggingface.co/katuni4ka/tiny-random-dbrx) on the Paladiso/dataset_91d2e9f8-99da-4b30-8acb-ee8b1520d1bc dataset.
+This model is a fine-tuned version of [openlm-research/open_llama_3b](https://huggingface.co/openlm-research/open_llama_3b) on the Paladiso/dataset_b98c9df0-5280-40e9-9125-37aafc6279e9 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 11.5
+- Loss: 0.9400
 
 ## Model description
 
@@ -137,9 +140,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 46.0          | 0.0001 | 3    | 11.5            |
-| 46.0          | 0.0002 | 6    | 11.5            |
-| 46.0          | 0.0003 | 9    | 11.5            |
+| 0.9502        | 0.0030 | 3    | 1.0558          |
+| 1.1032        | 0.0060 | 6    | 1.0392          |
+| 0.9992        | 0.0090 | 9    | 0.9400          |
 
 
 ### Framework versions