Remove internal validation details
Browse files
README.md
CHANGED
|
@@ -31,11 +31,6 @@ The mixed INT4 `.litertlm` artifact was produced with a TorchAO-based quantize-f
|
|
| 31 |
The mixed INT4 bundle also uses LiteRT-LM StableHLO composite ops for attention/cache execution, including `odml.runtime_bmm` and `odml.cache_update`.
|
| 32 |
|
| 33 |
|
| 34 |
-
## Validation
|
| 35 |
-
|
| 36 |
-
The mixed INT4 bundle was validated during conversion with `is_valid_tflite=True`, `unsupported_count=0`, and `contract_ok=True`.
|
| 37 |
-
|
| 38 |
-
|
| 39 |
## Performance
|
| 40 |
|
| 41 |
Desktop benchmark: AMD Radeon AI PRO R9700, LiteRT-LM WebGPU, 256 prefill tokens, 32 decode tokens. Android rows use LiteRT-LM v0.13.1 with GPU OpenCL, 256 prefill tokens, and 64 decode tokens. Values report the warmed iteration from a two-iteration run unless noted.
|
|
|
|
| 31 |
The mixed INT4 bundle also uses LiteRT-LM StableHLO composite ops for attention/cache execution, including `odml.runtime_bmm` and `odml.cache_update`.
|
| 32 |
|
| 33 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
## Performance
|
| 35 |
|
| 36 |
Desktop benchmark: AMD Radeon AI PRO R9700, LiteRT-LM WebGPU, 256 prefill tokens, 32 decode tokens. Android rows use LiteRT-LM v0.13.1 with GPU OpenCL, 256 prefill tokens, and 64 decode tokens. Values report the warmed iteration from a two-iteration run unless noted.
|