Text Generation
Transformers
llama
Inference Endpoints
bhenrym14 commited on
Commit
4fb4bad
1 Parent(s): 4170423

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +6 -6
README.md CHANGED
@@ -38,13 +38,13 @@ Here I explore whether training on long sequences that have clear conceptual dep
38
 
39
  ## Relative Performance (perplexity)
40
 
41
- | Context (tokens) | airophin-13b-pntk-16k-fp16| bhenrym14/airoboros-13b-gpt4-1.4.1-PI-8192-fp16 |bhenrym14/airoboros-33b-gpt4-1.4.1-lxctx-PI-16384-fp16 | jondurbin/airoboros-33B-gpt4-1.4-GPTQ |
42
  | ---| ----- | -----| ------| --- |
43
- | 512 | 7.62 | 8.24 | 7.90 | **6.36** |
44
- | 1024 | 6.20 | 6.71 | 6.17 | **5.12** |
45
- | 2048 | 5.38 | 5.87 | 5.23 | **4.43** |
46
- | 4096 | 5.08 | 5.50 | **4.91** | 54.5 |
47
- | 8192 | 4.90 | 5.32 | -- | -- |
48
  | 12000 | 4.82 | 56.1 | -- | -- |
49
 
50
  - This model is competitive with the Llama-1 33b variants, outperforming the best long context model for short sequences.
 
38
 
39
  ## Relative Performance (perplexity)
40
 
41
+ | Context (tokens) | airophin-13b-pntk-16k-fp16| bhenrym14/airoboros-13b-gpt4-1.4.1-PI-8192-fp16 |bhenrym14/airoboros-33b-gpt4-1.4.1-lxctx-PI-16384-fp16 | jondurbin/airoboros-l2-13b-gpt4-1.4.1 |
42
  | ---| ----- | -----| ------| --- |
43
+ | 512 | 7.62 | 8.24 | 7.90 | **7.23** |
44
+ | 1024 | 6.20 | 6.71 | 6.17 | **5.85** |
45
+ | 2048 | 5.38 | 5.87 | 5.23 | **5.07** |
46
+ | 4096 | 5.08 | 5.50 | **4.91** | 4.77 |
47
+ | 8192 | 4.90 | 5.32 | -- | 57.1 |
48
  | 12000 | 4.82 | 56.1 | -- | -- |
49
 
50
  - This model is competitive with the Llama-1 33b variants, outperforming the best long context model for short sequences.