Upload LlamaForCausalLM

README.md

@@ -6,10 +6,11 @@ tags:
 - finance
 pipeline_tag: text-generation
 widget:
-- text: "How do call options benefit the buyer?"
-  example_title: "Domain Specific QA"
-- text: "Why might a trader choose to quickly exit a losing position, even if they still believe in the original trade idea?"
-  example_title: "Not-Advice"
+- text: How do call options benefit the buyer?
+  example_title: Domain Specific QA
+- text: Why might a trader choose to quickly exit a losing position, even if they
+    still believe in the original trade idea?
+  example_title: Not-Advice
 inference:
   parameters:
     temperature: 0.3

config.json

@@ -1,4 +1,5 @@
 {
+  "_name_or_path": "models/irai/devoted-leaf-806-default-ft/merged",
   "architectures": [
     "LlamaForCausalLM"
   ],
@@ -10,7 +11,7 @@
   "hidden_size": 2048,
   "initializer_range": 0.02,
   "intermediate_size": 5632,
-  "max_position_embeddings": 2048,
+  "max_position_embeddings": 4096,
   "model_type": "llama",
   "num_attention_heads": 32,
   "num_hidden_layers": 22,
@@ -22,6 +23,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "float32",
   "transformers_version": "4.39.3",
-  "use_cache": true,
+  "use_cache": false,
   "vocab_size": 32000
 }

generation_config.json

@@ -1,5 +1,6 @@
 {
   "bos_token_id": 1,
+  "do_sample": true,
   "eos_token_id": 2,
   "max_length": 2048,
   "pad_token_id": 0,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b66b7f58d9f8f586b280d8bb56c5a657709458285f32dea05a1f49233f973dad
+oid sha256:6a5e0a13e3302318e44ce34263025c2da9db4eb4f8d646adaa85ace70e3deb45
 size 4400216536