first commit

config.json

@@ -1,4 +1,5 @@
 {
+  "_name_or_path": "/share/kuleshov/vk379/llamas/llama-7b-4bit",
   "architectures": [
     "LLaMAForCausalLM"
   ],
@@ -8,6 +9,7 @@
   "hidden_size": 4096,
   "initializer_range": 0.02,
   "intermediate_size": 11008,
+  "max_position_embeddings": 2048,
   "max_sequence_length": 2048,
   "model_type": "llama",
   "num_attention_heads": 32,
@@ -16,7 +18,7 @@
   "rms_norm_eps": 1e-06,
   "tie_word_embeddings": false,
   "torch_dtype": "float16",
-  "transformers_version": "4.27.0.dev0",
+  "transformers_version": "4.28.0",
   "use_cache": true,
   "vocab_size": 32000
 }

quant_config.json

@@ -0,0 +1,11 @@
+{
+  "dataset": "c4",
+  "bits": 4,
+  "nsamples": 128,
+  "groupsize": 64,
+  "act_order": true,
+  "percdamp": 0.01,
+  "seed": 0,
+  "nearest": false,
+  "save": "/tmp/llama-out-test"
+}

quantized_weights.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2be0456edba63d31f64be986886884607b76a0aae01b932f3fd4f0024f07d85b
+size 4020703149