Upload folder using huggingface_hub

README.md

@@ -0,0 +1,79 @@
+---
+base_model:
+- cstr/llama3.1-8b-spaetzle-v117
+- cstr/llama3.1-8b-spaetzle-v115
+- cstr/llama3.1-8b-spaetzle-v109
+- cstr/llama3.1-8b-spaetzle-v113
+tags:
+- merge
+- mergekit
+- lazymergekit
+- cstr/llama3.1-8b-spaetzle-v117
+- cstr/llama3.1-8b-spaetzle-v115
+- cstr/llama3.1-8b-spaetzle-v109
+- cstr/llama3.1-8b-spaetzle-v113
+---
+
+# llama3.1-8b-spaetzle-v119
+
+llama3.1-8b-spaetzle-v119 is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
+* [cstr/llama3.1-8b-spaetzle-v117](https://huggingface.co/cstr/llama3.1-8b-spaetzle-v117)
+* [cstr/llama3.1-8b-spaetzle-v115](https://huggingface.co/cstr/llama3.1-8b-spaetzle-v115)
+* [cstr/llama3.1-8b-spaetzle-v109](https://huggingface.co/cstr/llama3.1-8b-spaetzle-v109)
+* [cstr/llama3.1-8b-spaetzle-v113](https://huggingface.co/cstr/llama3.1-8b-spaetzle-v113)
+
+## 🧩 Configuration
+
+```yaml
+models:
+  - model: cstr/llama3.1-8b-spaetzle-v90
+    # no parameters necessary for base model
+  - model: cstr/llama3.1-8b-spaetzle-v117
+    parameters:
+      density: 0.65
+      weight: 0.2
+  - model: cstr/llama3.1-8b-spaetzle-v115
+    parameters:
+      density: 0.65
+      weight: 0.2
+  - model: cstr/llama3.1-8b-spaetzle-v109
+    parameters:
+      density: 0.65
+      weight: 0.25
+  - model: cstr/llama3.1-8b-spaetzle-v113
+    parameters:
+      density: 0.65
+      weight: 0.1
+merge_method: dare_ties
+base_model: cstr/llama3.1-8b-spaetzle-v90
+parameters:
+  int8_mask: true
+dtype: bfloat16
+random_seed: 0
+tokenizer_source: base
+```
+
+## 💻 Usage
+
+```python
+!pip install -qU transformers accelerate
+
+from transformers import AutoTokenizer
+import transformers
+import torch
+
+model = "cstr/llama3.1-8b-spaetzle-v119"
+messages = [{"role": "user", "content": "What is a large language model?"}]
+
+tokenizer = AutoTokenizer.from_pretrained(model)
+prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+pipeline = transformers.pipeline(
+    "text-generation",
+    model=model,
+    torch_dtype=torch.float16,
+    device_map="auto",
+)
+
+outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
+print(outputs[0]["generated_text"])
+```