End of training

README.md

@@ -18,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/phi-1_5](https://huggingface.co/microsoft/phi-1_5) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 2.1482
+- Loss: 0.3252
 
 ## Model description
 
@@ -37,8 +37,8 @@ More information needed
 
 The following `bitsandbytes` quantization config was used during training:
 - quant_method: bitsandbytes
-- load_in_8bit: False
-- load_in_4bit: True
+- _load_in_8bit: False
+- _load_in_4bit: True
 - llm_int8_threshold: 6.0
 - llm_int8_skip_modules: None
 - llm_int8_enable_fp32_cpu_offload: False
@@ -46,6 +46,8 @@ The following `bitsandbytes` quantization config was used during training:
 - bnb_4bit_quant_type: nf4
 - bnb_4bit_use_double_quant: True
 - bnb_4bit_compute_dtype: float16
+- load_in_4bit: True
+- load_in_8bit: False
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
@@ -55,7 +57,7 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: cosine
-- training_steps: 3
+- training_steps: 500
 
 ### Training results
 

runs/Jan31_13-01-38_5a17fd8cf618/events.out.tfevents.1706708824.5a17fd8cf618.346.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de325cb7286a7ff6c42851b8cc64fcbbdfd37bd84a2ba7542fcc072b0c0f890c
+size 359