Upload LlamaForCausalLM

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "unsloth/tinyllama",
+  "_name_or_path": "unsloth/tinyllama-bnb-4bit",
   "architectures": [
     "LlamaForCausalLM"
   ],
@@ -11,7 +11,7 @@
   "hidden_size": 2048,
   "initializer_range": 0.02,
   "intermediate_size": 5632,
-  "max_position_embeddings": 4096,
+  "max_position_embeddings": 2048,
   "model_type": "llama",
   "num_attention_heads": 32,
   "num_hidden_layers": 22,
@@ -19,7 +19,7 @@
   "pad_token_id": 0,
   "pretraining_tp": 1,
   "quantization_config": {
-    "bnb_4bit_compute_dtype": "float16",
+    "bnb_4bit_compute_dtype": "bfloat16",
     "bnb_4bit_quant_type": "nf4",
     "bnb_4bit_use_double_quant": true,
     "llm_int8_enable_fp32_cpu_offload": false,
@@ -31,15 +31,11 @@
     "quant_method": "bitsandbytes"
   },
   "rms_norm_eps": 1e-05,
-  "rope_scaling": {
-    "factor": 2.0,
-    "type": "linear"
-  },
+  "rope_scaling": null,
   "rope_theta": 10000.0,
   "tie_word_embeddings": false,
-  "torch_dtype": "float16",
-  "transformers_version": "4.37.0.dev0",
-  "unsloth_version": "2024.1",
+  "torch_dtype": "float32",
+  "transformers_version": "4.35.2",
   "use_cache": true,
   "vocab_size": 32000
 }

generation_config.json

@@ -3,5 +3,5 @@
   "eos_token_id": 2,
   "max_length": 2048,
   "pad_token_id": 0,
-  "transformers_version": "4.37.0.dev0"
+  "transformers_version": "4.35.2"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cef660c78bd8423780c49b876d55ad9fc312886228b40a3312ab9517bc42fbcc
-size 762453342
+oid sha256:60b943c6c08c9c855653f13225405ea8dd459424339385da2694deae9c382023
+size 4400216536