params : allow penalty_last_n == -1 to be equal to context size

ggml-ci

Author: ggerganov

common/arg.cpp

@@ -909,6 +909,9 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
         {"--repeat-last-n"}, "N",
         string_format("last n tokens to consider for penalize (default: %d, 0 = disabled, -1 = ctx_size)", params.sampling.penalty_last_n),
         [](common_params & params, int value) {
+            if (value < -1) {
+                throw std::runtime_error(string_format("error: invalid repeat-last-n = %d\n", value));
+            }
             params.sampling.penalty_last_n = value;
             params.sampling.n_prev = std::max(params.sampling.n_prev, params.sampling.penalty_last_n);
         }
@@ -963,6 +966,9 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
         {"--dry-penalty-last-n"}, "N",
         string_format("set DRY penalty for the last n tokens (default: %d, 0 = disable, -1 = context size)", params.sampling.dry_penalty_last_n),
         [](common_params & params, int value) {
+            if (value < -1) {
+                throw std::runtime_error(string_format("error: invalid dry-penalty-last-n = %d\n", value));
+            }
             params.sampling.dry_penalty_last_n = value;
         }
     ).set_sparam());

common/common.cpp

@@ -945,6 +945,16 @@ struct common_init_result common_init_from_params(common_params & params) {
         params.sampling.logit_bias.push_back({llama_token_eos(model), -INFINITY});
     }
 
+    if (params.sampling.penalty_last_n == -1) {
+        LOG_INF("%s: setting penalty_last_n to ctx_size = %d\n", __func__, llama_n_ctx(lctx));
+        params.sampling.penalty_last_n = llama_n_ctx(lctx);
+    }
+
+    if (params.sampling.dry_penalty_last_n == -1) {
+        LOG_INF("%s: setting dry_penalty_last_n to ctx_size = %d\n", __func__, llama_n_ctx(lctx));
+        params.sampling.dry_penalty_last_n = llama_n_ctx(lctx);
+    }
+
     if (params.warmup) {
         LOG_WRN("%s: warming up the model with an empty run - please wait ... (--no-warmup to disable)\n", __func__);
 

examples/server/server.cpp

@@ -183,6 +183,7 @@ struct server_task {
 
     static slot_params params_from_json_cmpl(
             const llama_model * model,
+            const llama_context * ctx,
             const common_params & params_base,
             const json & data) {
         slot_params params;
@@ -237,8 +238,27 @@ struct server_task {
         params.speculative.n_min = std::max(params.speculative.n_min, 2);
         params.speculative.n_max = std::max(params.speculative.n_max, 0);
 
+        // TODO: add more sanity checks for the input parameters
+
+        if (params.sampling.penalty_last_n < -1) {
+            throw std::runtime_error("Error: repeat_last_n must be >= -1");
+        }
+
+        if (params.sampling.dry_penalty_last_n < -1) {
+            throw std::runtime_error("Error: dry_penalty_last_n must be >= -1");
+        }
+
+        if (params.sampling.penalty_last_n == -1) {
+            // note: should be the slot's context and not the full context, but it's ok
+            params.sampling.penalty_last_n = llama_n_ctx(ctx);
+        }
+
+        if (params.sampling.dry_penalty_last_n == -1) {
+            params.sampling.dry_penalty_last_n = llama_n_ctx(ctx);
+        }
+
         if (params.sampling.dry_base < 1.0f) {
-           params.sampling.dry_base = defaults.sampling.dry_base;
+            params.sampling.dry_base = defaults.sampling.dry_base;
         }
 
         // sequence breakers for DRY
@@ -3379,7 +3399,7 @@ int main(int argc, char ** argv) {
                 task.index = i;
 
                 task.prompt_tokens    = std::move(tokenized_prompts[i]);
-                task.params           = server_task::params_from_json_cmpl(ctx_server.model, ctx_server.params_base, data);
+                task.params           = server_task::params_from_json_cmpl(ctx_server.model, ctx_server.ctx, ctx_server.params_base, data);
                 task.id_selected_slot = json_value(data, "id_slot", -1);
 
                 // OAI-compat