Training in progress, step 6000

README.md

@@ -27,7 +27,7 @@ print(output["generated_text"])
 
 ## Training procedure
 
-[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/g-puca1-deloitte/llmv3/runs/gfmzjww9) 
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/g-puca1-deloitte/llmv3/runs/16peb9vx) 
 
 
 This model was trained with SFT.

adapter_config.json

@@ -13,7 +13,7 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 32,
+  "lora_alpha": 16,
   "lora_bias": false,
   "lora_dropout": 0.0,
   "megatron_config": null,
@@ -21,14 +21,14 @@
   "modules_to_save": null,
   "peft_type": "LORA",
   "qalora_group_size": 16,
-  "r": 32,
+  "r": 16,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
     "q_proj",
+    "v_proj",
     "o_proj",
-    "k_proj",
-    "v_proj"
+    "k_proj"
   ],
   "target_parameters": null,
   "task_type": "CAUSAL_LM",

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3f48690fe905fa458df940f8c6d1cc50905f9e24c570a59712fc70375c24dbe9
-size 95590192
+oid sha256:2d4782a25ead7ec14e692a9d7ba6a1c622721b4598dea716bb0aab641c8b1002
+size 47814376

modelopt_state_train.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:67af8a09de4a37a23265511d219008b5b6d6ac9ffd6b23c82f9b083268177dd3
+oid sha256:543a2762eb7388bc5b2aad89dbe66a41728b3efab046c4145d3788188ed452e1
 size 975611

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:899a5a84d62dba2fc70f4f69f4c94485813c860169412e064f5134c673edbe8f
+oid sha256:96e67e32bc5b130bd9dc87c8487c8030dd801ddfefd04c3cf3dfdc3195b19e55
 size 6353