doberst commited on
Commit
694561f
·
verified ·
1 Parent(s): 362e961

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -3
README.md CHANGED
@@ -3,11 +3,11 @@ license: apache-2.0
3
  inference: false
4
  ---
5
 
6
- # bling-phi-3
7
 
8
  <!-- Provide a quick summary of what the model is/does. -->
9
 
10
- bling-phi-3 is part of the BLING ("Best Little Instruct No-GPU") model series, RAG-instruct trained on top of a Microsoft Phi-3.5 base model.
11
 
12
 
13
  ### Benchmark Tests
@@ -23,9 +23,13 @@ Evaluated against the benchmark test: [RAG-Instruct-Benchmark-Tester](https://
23
  --Summarization Quality (1-5): 4 (Above Average)
24
  --Hallucinations: No hallucinations observed in test runs.
25
 
 
 
 
 
26
  For test run results (and good indicator of target use cases), please see the files ("core_rag_test" and "answer_sheet" in this repo).
27
 
28
- Note: compare results with [bling-phi-2](https://www.huggingface.co/llmware/bling-phi-2-v0), and [dragon-mistral-7b](https://www.huggingface.co/llmware/dragon-mistral-7b-v0).
29
 
30
 
31
  ### Model Description
 
3
  inference: false
4
  ---
5
 
6
+ # bling-phi-3.5
7
 
8
  <!-- Provide a quick summary of what the model is/does. -->
9
 
10
+ bling-phi-3.5 is part of the BLING ("Best Little Instruct No-GPU") model series, RAG-instruct trained on top of a Microsoft Phi-3.5 base model.
11
 
12
 
13
  ### Benchmark Tests
 
23
  --Summarization Quality (1-5): 4 (Above Average)
24
  --Hallucinations: No hallucinations observed in test runs.
25
 
26
+ Note: test results were not produced with the pytorch packaging of the model, but rather the [gguf quantized version](https://www.huggingface.co/llmware/bling-phi-3.5-gguf).
27
+
28
+ It is possible that there will be minor variations with this fp16 pytorch model, which is released as a base for further fine-tuning and porting to other formats (ONNX, OpenVino), while the gguf quantized version is recommended for inference (on Mac and CUDA in particular).
29
+
30
  For test run results (and good indicator of target use cases), please see the files ("core_rag_test" and "answer_sheet" in this repo).
31
 
32
+ Note: compare results with [bling-phi-3](https://www.huggingface.co/llmware/bling-phi-3).
33
 
34
 
35
  ### Model Description