sean-m
/
kendryte-standalone-sdk
mirror of https://github.com/kendryte/kendryte-standalone-sdk.git
1
0
Fork 0
Code Issues Releases Wiki Activity

Kpu support fc for first layer

reformat
zhaozhongxiang 2019-05-31 13:38:55 +08:00
parent d5ca698b14
commit e2dd551e09
1 changed files with 49 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

60
lib/drivers/kpu.c
View File

@ -757,6 +757,27 @@ static void kpu_kmodel_input_with_padding(const kpu_layer_argument_t *layer, con
kpu_upload_core(width, height, channels, src, layer->image_addr.data.image_src_addr);
}
static void kpu_kmodel_input_float(const float* src, float* dest, size_t count)
{
memcpy(dest, src, count * sizeof(float));
}
static void kpu_float_activation(float *data, size_t count, kpu_model_activation_t act)
{
size_t i;
if (act == KLA_RELU)
{
for (i = 0; i < count; i++)
data[i] = max(data[i], 0);
}
else if (act == KLA_RELU6)
{
for (i = 0; i < count; i++)
data[i] = min(max(data[i], 0), 6);
}
}
static void kpu_kmodel_add(const kpu_model_add_layer_argument_t *arg, kpu_model_context_t *ctx)
{
const float *src_a = (const float *)(ctx->main_buffer + arg->main_mem_in_a_address);
@ -1195,6 +1216,8 @@ static void kpu_kmodel_fully_connected(const kpu_model_fully_connected_layer_arg
dest[oc] = sum + bias[oc];
}
}
kpu_float_activation(dest, out_channels, arg->act);
}
static void kpu_tf_flatten(const kpu_model_tf_flatten_layer_argument_t *arg, kpu_model_context_t *ctx)
@ -1657,19 +1680,34 @@ int kpu_run_kmodel(kpu_model_context_t *ctx, const uint8_t *src, dmac_channel_nu
plic_irq_enable(IRQN_AI_INTERRUPT);
const kpu_model_layer_header_t *first_layer_header = ctx->layer_headers;
if (first_layer_header->type != KL_K210_CONV)
return -1;
const kpu_model_conv_layer_argument_t *first_layer = (const kpu_model_conv_layer_argument_t *)ctx->body_start;
kpu_layer_argument_t layer_arg = *(volatile kpu_layer_argument_t *)(ctx->model_buffer + first_layer->layer_offset);
if ((layer_arg.image_size.data.i_row_wid + 1) % 64 != 0)
switch (first_layer_header->type)
{
kpu_kmodel_input_with_padding(&layer_arg, src);
ai_step_not_isr(ctx);
}
else
{
kpu_input_dma(&layer_arg, src, ctx->dma_ch, ai_step, ctx);
case KL_K210_CONV:
{
const kpu_model_conv_layer_argument_t *first_layer = (const kpu_model_conv_layer_argument_t *)ctx->body_start;
kpu_layer_argument_t layer_arg = *(volatile kpu_layer_argument_t *)(ctx->model_buffer + first_layer->layer_offset);
if ((layer_arg.image_size.data.i_row_wid + 1) % 64 != 0)
{
kpu_kmodel_input_with_padding(&layer_arg, src);
ai_step_not_isr(ctx);
}
else
{
kpu_input_dma(&layer_arg, src, ctx->dma_ch, ai_step, ctx);
}
}
break;
case KL_FULLY_CONNECTED:
{
const kpu_model_fully_connected_layer_argument_t *first_layer = (const kpu_model_fully_connected_layer_argument_t *)ctx->body_start;
kpu_kmodel_input_float((const float *)src, (float *)(ctx->main_buffer + first_layer->main_mem_in_address), first_layer->in_channels);
ai_step_not_isr(ctx);
}
break;
default:
return -1;
}
return 0;