End of training

README.md

@@ -72,9 +72,9 @@ lora_target_modules:
 # hub_model_id: dyang415/mixtral-lora-v0
 
 
-gradient_accumulation_steps: 2
-micro_batch_size: 1
-num_epochs: 1
+gradient_accumulation_steps: 4
+micro_batch_size: 2
+num_epochs: 4
 optimizer: paged_adamw_8bit
 lr_scheduler: cosine
 learning_rate: 0.0002
@@ -141,18 +141,18 @@ The following `bitsandbytes` quantization config was used during training:
 
 The following hyperparameters were used during training:
 - learning_rate: 0.0002
-- train_batch_size: 1
-- eval_batch_size: 1
+- train_batch_size: 2
+- eval_batch_size: 2
 - seed: 42
 - distributed_type: multi-GPU
 - num_devices: 2
-- gradient_accumulation_steps: 2
-- total_train_batch_size: 4
-- total_eval_batch_size: 2
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 16
+- total_eval_batch_size: 4
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_steps: 10
-- num_epochs: 1
+- num_epochs: 4
 
 ### Training results
 

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:26c3ba2e76a2502bf3818d54d1b083a30e1ac7e46caafcca1cd55d56286d652d
+oid sha256:a19bf003a45ac34d209f815b8adc7dd18c65636db4c13c6e9e76d402c884a079
 size 109144269

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:27728d1b7c09b9df9214cf0d980dff640c42cded1e8ece183959acd96c00a361
+oid sha256:366bca99dab2f58d415e34eaa2b31474510e19c78b5a52a6e16dd6f23ec633bc
 size 109086416

runs/Mar02_18-57-35_azure-jap/events.out.tfevents.1709405856.azure-jap.71737.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8a70e6979f76ac31aa710be6e99d3244b112a22ac4bad1a8cd1f67821d10147a
-size 50355
+oid sha256:883d8b3744cd16896e57e715f490f13c01059301b6007724c410622bcf291b04
+size 65781