Dracones
/

Merged-RP-Stew-V2-34B_exl2_5.0bpw

Text Generation

Not-For-All-Audiences

text-generation-inference

Inference Endpoints

Model card Files Files and versions Community

Dracones commited on Apr 6

Commit

989ae70

•

1 Parent(s): c305ea5

Upload README.md with huggingface_hub

Files changed (1) hide show

README.md +10 -10

README.md CHANGED Viewed

@@ -41,36 +41,36 @@ Below are the perplexity scores for the EXL2 models. A lower score is better.
 ## EQ Bench
-Here are the EQ Bench scores for the EXL2 quants using Alpaca, ChatML and Vicuna-Chat prompt templates. A higher score is better.
 | Quant Size | Instruct Template | Score |
 |------------|-------------------|-------|
 | 8.0        | Alpaca            | 75.02 |
 | 8.0        | ChatML            | 73.82 |
-| 8.0        | Vicuna-Chat       | 73.53 |
 | 7.0        | Alpaca            | 74.96 |
 | 7.0        | ChatML            | 74.22 |
-| 7.0        | Vicuna-Chat       | 72.71 |
 | 6.0        | Alpaca            | 75.11 |
 | 6.0        | ChatML            | 74.29 |
-| 6.0        | Vicuna-Chat       | 74.16 |
 | 5.0        | Alpaca            | 74.11 |
 | 5.0        | ChatML            | 72.74 |
-| 5.0        | Vicuna-Chat       | 73.28 |
 | 4.5        | Alpaca            | 73.18 |
 | 4.5        | ChatML            | 73.62 |
-| 4.5        | Vicuna-Chat       | 73.32 |
 | 4.0        | Alpaca            | 71.85 |
 | 4.0        | ChatML            | 73.99 |
-| 4.0        | Vicuna-Chat       | 74.74 |
 | 3.5        | Alpaca            | 73.73 |
 | 3.5        | ChatML            | 72.07 |
-| 3.5        | Vicuna-Chat       | 73.56 |
-### Vicuna Chat Template
-This is the Vicuna-Chat template yaml that was used in EQ bench. It was tested in Text Generation Web UI and seemed to produce accurate results.
 ```yaml
 user: "USER:"

 ## EQ Bench
+Here are the EQ Bench scores for the EXL2 quants using Alpaca, ChatML and Chat-Vicuna prompt templates. A higher score is better.
 | Quant Size | Instruct Template | Score |
 |------------|-------------------|-------|
 | 8.0        | Alpaca            | 75.02 |
 | 8.0        | ChatML            | 73.82 |
+| 8.0        | Chat-Vicuna       | 73.53 |
 | 7.0        | Alpaca            | 74.96 |
 | 7.0        | ChatML            | 74.22 |
+| 7.0        | Chat-Vicuna       | 72.71 |
 | 6.0        | Alpaca            | 75.11 |
 | 6.0        | ChatML            | 74.29 |
+| 6.0        | Chat-Vicuna       | 74.16 |
 | 5.0        | Alpaca            | 74.11 |
 | 5.0        | ChatML            | 72.74 |
+| 5.0        | Chat-Vicuna       | 73.28 |
 | 4.5        | Alpaca            | 73.18 |
 | 4.5        | ChatML            | 73.62 |
+| 4.5        | Chat-Vicuna       | 73.32 |
 | 4.0        | Alpaca            | 71.85 |
 | 4.0        | ChatML            | 73.99 |
+| 4.0        | Chat-Vicuna       | 74.74 |
 | 3.5        | Alpaca            | 73.73 |
 | 3.5        | ChatML            | 72.07 |
+| 3.5        | Chat-Vicuna       | 73.56 |
+### Chat-Vicuna Template
+This is the Chat-Vicuna template yaml that was used in EQ bench. It was tested in Text Generation Web UI and seemed to produce accurate results.
 ```yaml
 user: "USER:"