End of training

README.md

@@ -71,16 +71,17 @@ pad_to_sequence_len: true
 resume_from_checkpoint: null
 s2_attention: null
 sample_packing: false
-save_steps: 1
+save_steps: 5
 save_strategy: steps
 sequence_len: 4096
 special_tokens:
-  pad_token: <|end_of_text|>
+  pad_token: ' '
 strict: false
 tf32: false
 tokenizer_type: AutoTokenizer
 train_on_inputs: false
 val_set_size: 0.05
+wandb_mode: disabled
 warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
@@ -93,7 +94,7 @@ xformers_attention: null
 
 This model is a fine-tuned version of [unsloth/Llama-3.2-3B-Instruct](https://huggingface.co/unsloth/Llama-3.2-3B-Instruct) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5293
+- Loss: 5.0689
 
 ## Model description
 
@@ -127,14 +128,14 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 1.3547        | 0.8   | 1    | 0.5278          |
-| 1.2703        | 1.6   | 2    | 0.5293          |
+| 5.4988        | 0.8   | 1    | 5.0751          |
+| 5.2725        | 1.6   | 2    | 5.0689          |
 
 
 ### Framework versions
 
 - PEFT 0.13.0
-- Transformers 4.45.0
+- Transformers 4.45.1
 - Pytorch 2.3.1+cu121
 - Datasets 2.21.0
 - Tokenizers 0.20.0

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aeb914d6e9994726a20c95dbf08ff9cff38dc401127d09960189c23ca6a0ae86
-size 194652458
+oid sha256:f63cc530a8bbed71ec5c9e53b4a02a21d00fd2376226c3f48343ec58132cbbe5
+size 982663982