Upload 4 files

README.md

@@ -0,0 +1,89 @@
+---
+datasets:
+- tatsu-lab/alpaca
+language:
+- en
+---
+### Model card for Alpaca-30B
+
+This is a Llama model instruction-finetuned with LoRa for 3 epochs on the Tatsu Labs Alpaca dataset. It was trained in 8bit mode.
+
+To run this model, you can run the following or use the following repo for [generation](https://github.com/aspctu/alpaca-lora). 
+
+```
+# Code adapted from https://github.com/tloen/alpaca-lora
+import torch
+from peft import PeftModel
+import transformers
+
+from transformers import LlamaTokenizer, LlamaForCausalLM, GenerationConfig
+
+tokenizer = LlamaTokenizer.from_pretrained("decapoda-research/llama-30b-hf")
+
+model = LlamaForCausalLM.from_pretrained(
+    "decapoda-research/llama-30b-hf",
+    load_in_8bit=True,
+    torch_dtype=torch.float16,
+    device_map="auto",
+)
+
+model = PeftModel.from_pretrained(
+    model, 
+    "baseten/alpaca-30b",
+    torch_dtype=torch.float16
+)
+
+def generate_prompt(instruction, input=None):
+    if input:
+        return f"""Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
+
+### Instruction:
+{instruction}
+
+### Input:
+{input}
+
+### Response:"""
+    else:
+        return f"""Below is an instruction that describes a task. Write a response that appropriately completes the request.
+
+### Instruction:
+{instruction}
+
+### Response:"""
+
+
+model.eval()
+
+
+def evaluate(
+        instruction,
+        input=None,
+        temperature=0.1,
+        top_p=0.75,
+        top_k=40,
+        num_beams=4,
+        **kwargs,
+):
+    prompt = generate_prompt(instruction, input)
+    inputs = tokenizer(prompt, return_tensors="pt")
+    input_ids = inputs["input_ids"].to(device)
+    generation_config = GenerationConfig(
+        temperature=temperature,
+        top_p=top_p,
+        top_k=top_k,
+        num_beams=num_beams,
+        **kwargs,
+    )
+    with torch.no_grad():
+        generation_output = model.generate(
+            input_ids=input_ids,
+            generation_config=generation_config,
+            return_dict_in_generate=True,
+            output_scores=True,
+            max_new_tokens=2048,
+        )
+    s = generation_output.sequences[0]
+    output = tokenizer.decode(s)
+    return output.split("### Response:")[1].strip()
+```

adapter_config.json

@@ -0,0 +1,18 @@
+{
+  "base_model_name_or_path": "decapoda-research/llama-30b-hf",
+  "bias": "none",
+  "enable_lora": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "merge_weights": false,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0365143b9adc0323d00b43abbebcb6412e2a15397de20c8d1da45d96f950267f
+size 133