Training in progress, step 78

adapter_config.json

@@ -1,6 +1,9 @@
 {
   "alpha_pattern": {},
-  "auto_mapping": null,
+  "auto_mapping": {
+    "base_model_class": "LlamaForCausalLM",
+    "parent_library": "transformers.models.llama.modeling_llama"
+  },
   "base_model_name_or_path": "deepseek-ai/deepseek-coder-1.3b-base",
   "bias": "none",
   "fan_in_fan_out": false,
@@ -21,14 +24,14 @@
   "revision": null,
   "target_modules": [
     "q_proj",
-    "gate_proj",
-    "up_proj",
-    "down_proj",
     "k_proj",
     "v_proj",
-    "o_proj"
+    "o_proj",
+    "down_proj",
+    "up_proj",
+    "gate_proj"
   ],
-  "task_type": "CAUSAL_LM",
+  "task_type": null,
   "use_dora": false,
   "use_rslora": false
 }

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:df4981a5d99183ddc19ea0be289ddd85574431d785bc60cd1af939d5ab5468bc
+oid sha256:df9eb2ed74b1e675d615105e9605bbfc6c51e1b9fa8e05029530ec146cba3103
 size 119975992

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2f557196d72a559a8fd6e84831b7898f6194f8289cba766bb177db73f55fd75a
+oid sha256:6242fd1fb2c0752ba17fc2fd6da8da7820830ff64ad4cf3bab8bd9cdaaa1cb4a
 size 5560