x/imagegen: add naive TeaCache and FP8 quantization support (#13683)

TeaCache:
- Timestep embedding similarity caching for diffusion models
- Polynomial rescaling with configurable thresholds
- Reduces transformer forward passes by ~30-50%

FP8 quantization:
- Support for FP8 quantized models (8-bit weights with scales)
- QuantizedMatmul on Metal, Dequantize on CUDA
- Client-side quantization via ollama create --quantize fp8

Other bug fixes:
- Fix `/api/show` API for image generation models
- Server properly returns model info (architecture, parameters, quantization)
- Memory allocation optimizations
- CLI improvements for image generation

api/client.go

@@ -165,7 +165,7 @@ func (c *Client) do(ctx context.Context, method, path string, reqData, respData
 	return nil
 }
 
-const maxBufferSize = 512 * format.KiloByte
+const maxBufferSize = 8 * format.MegaByte
 
 func (c *Client) stream(ctx context.Context, method, path string, data any, fn func([]byte) error) error {
 	var buf io.Reader