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: EleutherAI/pythia-410m-deduped
 tags:
 - axolotl
 - generated_from_trainer
 datasets:
-- Aivesa/dataset_7c40032f-e667-40ad-9658-3748512bf15b
+- Aivesa/dataset_f5aaf6b6-8b8a-4d62-a003-0a4288146cc5
 model-index:
-- name: 01290299-6ff6-484d-800c-9fc02709045a
+- name: a5f1d39f-30d4-4ab2-93d1-269d93a6989b
   results: []
 ---
 
@@ -21,17 +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: EleutherAI/pythia-410m-deduped
 bf16: auto
 chat_template: llama3
 dataset_prepared_path: /workspace/axolotl/data/prepared
 datasets:
 - ds_type: json
   format: custom
-  path: Aivesa/dataset_7c40032f-e667-40ad-9658-3748512bf15b
+  path: Aivesa/dataset_f5aaf6b6-8b8a-4d62-a003-0a4288146cc5
   type:
-    field_instruction: sentence1
-    field_output: sentence2
+    field_input: title
+    field_instruction: query
+    field_output: positive
     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: Aivesa/01290299-6ff6-484d-800c-9fc02709045a
+hub_model_id: Aivesa/a5f1d39f-30d4-4ab2-93d1-269d93a6989b
 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: <|endoftext|>
 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: 7c40032f-e667-40ad-9658-3748512bf15b
+wandb_name: f5aaf6b6-8b8a-4d62-a003-0a4288146cc5
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
-wandb_runid: 7c40032f-e667-40ad-9658-3748512bf15b
+wandb_runid: f5aaf6b6-8b8a-4d62-a003-0a4288146cc5
 warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
@@ -99,11 +102,11 @@ xformers_attention: null
 
 </details><br>
 
-# 01290299-6ff6-484d-800c-9fc02709045a
+# a5f1d39f-30d4-4ab2-93d1-269d93a6989b
 
-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_7c40032f-e667-40ad-9658-3748512bf15b dataset.
+This model is a fine-tuned version of [EleutherAI/pythia-410m-deduped](https://huggingface.co/EleutherAI/pythia-410m-deduped) on the Aivesa/dataset_f5aaf6b6-8b8a-4d62-a003-0a4288146cc5 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 3.1468
+- Loss: 3.4159
 
 ## Model description
 
@@ -137,9 +140,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 3.7746        | 0.0030 | 3    | 3.5979          |
-| 2.9603        | 0.0060 | 6    | 3.4694          |
-| 3.3668        | 0.0090 | 9    | 3.1468          |
+| 13.5865       | 0.0003 | 3    | 3.4376          |
+| 13.3023       | 0.0006 | 6    | 3.4332          |
+| 13.8277       | 0.0008 | 9    | 3.4159          |
 
 
 ### Framework versions