Update README.md

README.md

@@ -1,125 +1,67 @@
 ---
-base_model:
-- abacaj/phi-2-super
-tags:
-- mergekit
-- merge
-
+license: mit
+language:
+- en
+thumbnail: "https://cdn-uploads.huggingface.co/production/uploads/6589d7e6586088fd2784a12c/TqnMpteVAyfiiNHx4lVkU.png"
 ---
-# Teldrassil
+# You are welcome here, traveler.
+![image/png](https://cdn-uploads.huggingface.co/production/uploads/6589d7e6586088fd2784a12c/TqnMpteVAyfiiNHx4lVkU.png)
 
-This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
+### Named after the method used to create it, interleaving the layers of its predecessor to become far larger, giving it much more potential.
 
-## Merge Details
-### Merge Method
 
-This model was merged using the passthrough merge method.
+[Elothir](https://wowpedia.fandom.com/wiki/Elothir) was an ancient treeant, and I couldn't think of a better name for a model that was created using the passthrough method.
 
-### Models Merged
+The passthrough method differs significantly from the previous ones. By concatenating layers from different LLMs, it can produce models with an exotic number of parameters (e.g., 9B with two 7B parameter models). These models are often referred to as "frankenmerges" or "Frankenstein models" by the community.
 
-The following models were included in the merge:
-* [abacaj/phi-2-super](https://huggingface.co/abacaj/phi-2-super)
 
-### Configuration
+Many thanks to [Abacaj](https://huggingface.co/abacaj) for providing the [fine tuned weights](https://huggingface.co/abacaj/phi-2-super) that were used in the creation of this base model. You can find the full script for how the model was merged [here](https://huggingface.co/Replete-AI/Phi-Elothir/blob/main/mergekit_config.yml)...thanks to [KatyTheCutie](https://huggingface.co/KatyTheCutie) for helping me figure out how to make the model as big as I possibly could.
 
-The following YAML configuration was used to produce this model:
+## This idea was brought to me by [The Face of Goonery](https://huggingface.co/The-Face-Of-Goonery), also known as Caleb Morgan. I have him to thank if fine-tuning this model turns out to be a success
+# How to run inference:
 
-```yaml
-dtype: float16
-merge_method: passthrough
-slices:
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [0,2]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [1,3]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [2,4]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [3,5]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [4,6]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [5,7]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [6,8]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [7,9]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [8,10]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [9,11]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [10,12]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [11,13]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [12,14]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [13,15]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [14,16]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [15,17]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [16,18]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [17,19]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [18,20]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [19,21]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [20,22]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [21,23]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [22,24]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [23,25]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [24,26]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [25,27]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [26,28]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [27,29]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [28,30]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [29,31]
-  - sources:
-    - model: abacaj/phi-2-super
-      layer_range: [30,32]
+```python
+import transformers
+import torch
 
+if __name__ == "__main__":
+  model_name = "abacaj/phi-2-super"
+  tokenizer = transformers.AutoTokenizer.from_pretrained(model_name)
+  
+  model = (
+      transformers.AutoModelForCausalLM.from_pretrained(
+          model_name,
+      )
+      .to("cuda:0")
+      .eval()
+  )
+  
+  messages = [
+      {"role": "user", "content": "Hello, who are you?"}
+  ]
+  inputs = tokenizer.apply_chat_template(messages, return_tensors="pt").to(model.device)
+  input_ids_cutoff = inputs.size(dim=1)
+  
+  with torch.no_grad():
+      generated_ids = model.generate(
+          input_ids=inputs,
+          use_cache=True,
+          max_new_tokens=512,
+          temperature=0.2,
+          top_p=0.95,
+          do_sample=True,
+          eos_token_id=tokenizer.eos_token_id,
+          pad_token_id=tokenizer.pad_token_id,
+      )
+  
+  completion = tokenizer.decode(
+      generated_ids[0][input_ids_cutoff:],
+      skip_special_tokens=True,
+  )
+  
+  print(completion)
 ```
+
+# Chat template
+
+The model uses the same chat template as found in Mistral instruct models: