falcon-h1: fix wrong size param

younesbelkada · younesbelkada · commit 6dde986aef21 · 2025-07-09T11:38:36.000+04:00
diff --git a/src/llama-model.cpp b/src/llama-model.cpp
@@ -4557,7 +4557,7 @@ bool llama_model::load_tensors(llama_model_loader & ml) {
                     
                     // if output is NULL, init from the input tok embed
                     if (output == NULL) {
-                        output = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {n_embd, n_vocab}, TENSOR_DUPLICATED);
+                        output = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {hidden_size, n_vocab}, TENSOR_DUPLICATED);
                     }
 
                     for (int i = 0; i < n_layer; ++i) {

Original file line number	Diff line number	Diff line change
`@@ -4557,7 +4557,7 @@ bool llama_model::load_tensors(llama_model_loader & ml) {`
`4557`	`4557`
`4558`	`4558`	`// if output is NULL, init from the input tok embed`
`4559`	`4559`	`if (output == NULL) {`
`4560`		`- output = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {n_embd, n_vocab}, TENSOR_DUPLICATED);`
	`4560`	`+ output = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {hidden_size, n_vocab}, TENSOR_DUPLICATED);`
`4561`	`4561`	`}`
`4562`	`4562`
`4563`	`4563`	`for (int i = 0; i < n_layer; ++i) {`