Upload fine-tuned model and GGUF file from training session.

README.md

@@ -1,6 +1,7 @@
 ---
 base_model: HuggingFaceTB/SmolLM2-135M-Instruct
 library_name: peft
+pipeline_tag: text-generation
 tags:
 - base_model:adapter:HuggingFaceTB/SmolLM2-135M-Instruct
 - lora

adapter_config.json

@@ -1,11 +1,8 @@
 {
   "alpha_pattern": {},
-  "auto_mapping": {
-    "base_model_class": "LlamaForCausalLM",
-    "parent_library": "transformers.models.llama.modeling_llama"
-  },
+  "auto_mapping": null,
   "base_model_name_or_path": "HuggingFaceTB/SmolLM2-135M-Instruct",
-  "bias": "lora_only",
+  "bias": "none",
   "corda_config": null,
   "eva_config": null,
   "exclude_modules": null,
@@ -16,33 +13,28 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 128,
+  "lora_alpha": 64,
   "lora_bias": false,
-  "lora_dropout": 0.05,
+  "lora_dropout": 0.1,
   "megatron_config": null,
   "megatron_core": "megatron.core",
-  "modules_to_save": [
-    "embed_tokens",
-    "lm_head"
-  ],
+  "modules_to_save": null,
   "peft_type": "LORA",
   "qalora_group_size": 16,
-  "r": 64,
+  "r": 32,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "down_proj",
-    "o_proj",
+    "v_proj",
     "k_proj",
+    "down_proj",
     "gate_proj",
+    "o_proj",
     "up_proj",
-    "q_proj",
-    "v_proj",
-    "lm_head",
-    "embed_tokens"
+    "q_proj"
   ],
   "target_parameters": null,
-  "task_type": null,
+  "task_type": "CAUSAL_LM",
   "trainable_token_indices": null,
   "use_dora": false,
   "use_qalora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7770efca6d58143bc19b141efc17ce2e678b48a4956c59851609565f05a58a72
-size 191453608
+oid sha256:698ac3b5513872988146c3ec55b1e495ded6a961bf3434666d435829fa6ce6f9
+size 39131224

hyperlane-finetuned.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fbaacca02319051cc146c2812e4821dedf89a70723d3d5605e7706e0ce429964
-size 327508960
+oid sha256:ea303f7d820a0e962151aed9d4dac7e617686eea52a98845ae7736fa89d0993a
+size 270885792

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:41da9b094c3fcb1ae7c107637c01b541bc6033e7341038d8930048a5c242a97a
+oid sha256:fd8614a13250ac7bc9647cdb8d961d7e96383e9af4f08565e1f446ee9cc4480f
 size 6097