update model card
Browse files
README.md
CHANGED
|
@@ -27,31 +27,8 @@ Parameter counts are calculated from the tensors stored in the upstream checkpoi
|
|
| 27 |
|
| 28 |
### Performance Metrics
|
| 29 |
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
|
| 33 |
-
|---|---|
|
| 34 |
-
| Platform | Matrix6P |
|
| 35 |
-
| Data type | W8A8 |
|
| 36 |
-
| ViT image size | 448 × 448 |
|
| 37 |
-
| Sequence length | 512 |
|
| 38 |
-
| Maximum context length | 1024 |
|
| 39 |
-
| BPU cores (ViT / Prefill / Decode) | 4 / 4 / 4 |
|
| 40 |
-
|
| 41 |
-
#### Performance Results
|
| 42 |
-
|
| 43 |
-
| Metric | Value |
|
| 44 |
-
|---|---:|
|
| 45 |
-
| ViT latency | 41.535 ms |
|
| 46 |
-
| Time to first token (TTFT) | 84.448 ms |
|
| 47 |
-
| Prefill throughput | 13,265.273 tokens/s |
|
| 48 |
-
| Decode throughput | 70.473 tokens/s |
|
| 49 |
-
|
| 50 |
-
#### Memory Usage
|
| 51 |
-
|
| 52 |
-
| Metric | Value |
|
| 53 |
-
|---|---:|
|
| 54 |
-
| BPU memory | 2.4 GB |
|
| 55 |
-
| CPU memory | 0.79 GB |
|
| 56 |
|
| 57 |
> **Note:** TTFT includes preprocessing and ViT latency. Memory values represent the peak memory usage measured during the specified performance test.
|
|
|
|
| 27 |
|
| 28 |
### Performance Metrics
|
| 29 |
|
| 30 |
+
| Chips | Data Type | ViT Image Size | Sequence Length (tokens) | Maximum Context Length (tokens) | BPU Cores (ViT / Prefill / Decode) | ViT Latency (ms) | TTFT (ms) | Prefill TPS (token/s) | Decode TPS (token/s) | BPU Memory (GB) | CPU Memory (GB) |
|
| 31 |
+
|---|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
|
| 32 |
+
| Matrix6P | W8A8 | 448 × 448 | 512 | 1024 | 4 / 4 / 4 | 41.535 | 84.448 | 13,265.273 | 70.473 | 2.4 | 0.79 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
|
| 34 |
> **Note:** TTFT includes preprocessing and ViT latency. Memory values represent the peak memory usage measured during the specified performance test.
|