add gpu optimized wi4b32_wi8_afp32 model
#10
by CarreyZhan - opened
- README.md +1 -0
- minicpm_wi4b32_wi8_afp32_gpu_opt.litertlm +3 -0
README.md
CHANGED
|
@@ -25,6 +25,7 @@ This repository hosts the [**LiteRT-LM**](https://ai.google.dev/edge/litert-lm)
|
|
| 25 |
## Available Models
|
| 26 |
|
| 27 |
* **`minicpm_wi4b32_wi8_afp32.litertlm`**: This model features mixed INT4-block32(linear)/INT8(embed and lmhead) quantization (wi4b32_wi8) with FP32 activations (afp32).
|
|
|
|
| 28 |
* **`MiniCPM5-1B_dynamic_wi8_afp32.litertlm`**: This model features dynamic weight-only INT8 quantization (wi8) with FP32 activations (afp32). It is configured with static prefill memory allocation.
|
| 29 |
|
| 30 |
## Quantization Benchmark
|
|
|
|
| 25 |
## Available Models
|
| 26 |
|
| 27 |
* **`minicpm_wi4b32_wi8_afp32.litertlm`**: This model features mixed INT4-block32(linear)/INT8(embed and lmhead) quantization (wi4b32_wi8) with FP32 activations (afp32).
|
| 28 |
+
* **`minicpm_wi4b32_wi8_afp32_gpu_opt.litertlm`**: This model uses same quantization recipe with minicpm_wi4b32_wi8_afp32 but optimized for GPU execution.
|
| 29 |
* **`MiniCPM5-1B_dynamic_wi8_afp32.litertlm`**: This model features dynamic weight-only INT8 quantization (wi8) with FP32 activations (afp32). It is configured with static prefill memory allocation.
|
| 30 |
|
| 31 |
## Quantization Benchmark
|
minicpm_wi4b32_wi8_afp32_gpu_opt.litertlm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4e15a7cc735ae36e9888d341d44bd53c0579153d7a3379aaf958620eea4816e3
|
| 3 |
+
size 793034752
|