Add version endpoint and update create

NAMESPACE

@@ -31,6 +31,7 @@ export(test_connection)
 export(validate_message)
 export(validate_messages)
 export(validate_options)
+export(ver)
 importFrom(crayon,green)
 importFrom(crayon,red)
 importFrom(glue,glue)

NEWS.md

@@ -1,5 +1,8 @@
 # ollamar (development version)
 
+- Add `ver()` function to [retrieve Ollama version](https://github.com/ollama/ollama/blob/main/docs/api.md#version).
+- Update `create()` function.
+
 # ollamar 1.2.2
 
 - `generate()` and `chat()` support [structured output](https://ollama.com/blog/structured-outputs) via `format` parameter.

R/ollama.R

@@ -289,14 +289,14 @@ chat <- function(model, messages, tools = list(), stream = FALSE, format = list(
 
 
 
-#' Create a model from a Modelfile
+#' Create a model
 #'
-#' It is recommended to set `modelfile` to the content of the Modelfile rather than just set path.
+#' Create a model from another model, a safetensors directory (not implemented), or a GGUF file (not implemented).
 #'
-#' @param name Name of the model to create.
-#' @param modelfile Contents of the Modelfile as character string. Default is NULL.
+#' @param model Name of the model to create.
+#' @param from Name of an existing model to create the new model from.
+#' @param system System prompt for the model. Default is NULL.
 #' @param stream Enable response streaming. Default is FALSE.
-#' @param path The path to the Modelfile. Default is NULL.
 #' @param endpoint The endpoint to create the model. Default is "/api/create".
 #' @param host The base URL to use. Default is NULL, which uses Ollama's default base URL.
 #'
@@ -307,34 +307,22 @@ chat <- function(model, messages, tools = list(), stream = FALSE, format = list(
 #' @export
 #'
 #' @examplesIf test_connection(logical = TRUE)
-#' create("mario", "FROM llama3\nSYSTEM You are mario from Super Mario Bros.")
+#' create("mario", "deepseek-r1:1.5b", system = "You are Mario from Super Mario Bros.")
+#' model_avail("mario")  # check mario model has been created
+#' list_models()  # mario model has been created
 #' generate("mario", "who are you?", output = "text")  # model should say it's Mario
 #' delete("mario")  # delete the model created above
-create <- function(name, modelfile = NULL, stream = FALSE, path = NULL, endpoint = "/api/create", host = NULL) {
-
-    if (is.null(modelfile) && is.null(path)) {
-        stop("Either modelfile or path must be provided. Using modelfile is recommended.")
-    }
-
-    if (!is.null(modelfile) && !is.null(path)) {
-        stop("Only one of modelfile or path should be provided.")
-    }
-
-    if (!is.null(path)) {
-        if (file.exists(path)) {
-            modelfile <- paste0(readLines("inst/extdata/example_modefile.txt", warn = FALSE), collapse = "\n")
-            cat(paste0("Modefile\n", modelfile, "\n"))
-        } else {
-            stop("The path provided does not exist.")
-        }
-    }
+#' model_avail("mario")  # model no longer exists
+create <- function(model, from, system = NULL, stream = FALSE, endpoint = "/api/create", host = NULL) {
 
     req <- create_request(endpoint, host)
     req <- httr2::req_method(req, "POST")
 
+    # TODO: add other parameters
     body_json <- list(
-        name = name,
-        modelfile = modelfile,
+        model = model,
+        from = from,
+        system = system,
         stream = stream
     )
 
@@ -896,6 +884,40 @@ ps <- function(output = c("df", "resp", "jsonlist", "raw", "text"), endpoint = "
 
 
 
+#' Retrieve Ollama version
+#'
+#' @param endpoint The endpoint to list the running models. Default is "/api/version".
+#' @param host The base URL to use. Default is NULL, which uses Ollama's default base URL.
+#'
+#' @references
+#' [API documentation](https://github.com/ollama/ollama/blob/main/docs/api.md#version)
+#'
+#' @return A character string of the Ollama version.
+#' @export
+#'
+#' @examplesIf test_connection(logical = TRUE)
+#' ver()
+ver <- function(endpoint = "/api/version", host = NULL) {
+    req <- create_request(endpoint, host)
+    req <- httr2::req_method(req, "GET")
+    tryCatch(
+        {
+            resp <- httr2::req_perform(req)
+            return(resp_process(resp = resp, output = "text"))
+        },
+        error = function(e) {
+            stop(e)
+        }
+    )
+}
+
+
+
+
+
+
+
+
 
 
 

R/utils.R

@@ -275,6 +275,8 @@ resp_process <- function(resp, output = c("df", "jsonlist", "raw", "resp", "text
         } else if (output == "text") {
             return(df_response$name)
         }
+    } else if (grepl("api/version", resp$url)) {
+        return(httr2::resp_body_json(resp)$version)
     }
 }
 

_pkgdown.yml

@@ -45,6 +45,7 @@ reference:
       - embed
       - embeddings
       - ps
+      - ver
 
   - subtitle: API helpers
     desc: Work with and extend the Ollama API.

tests/testthat/test-chat.R

@@ -262,7 +262,7 @@ test_that("structured output", {
     )
 
     msg <- create_message("tell me about canada")
-    resp <- chat("llama3.1", msg, format = format)
+    resp <- chat("llama3.1:8b", msg, format = format)
     # content <- httr2::resp_body_json(resp)$message$content
     structured_output <- resp_process(resp, "structured")
     expect_equal(tolower(structured_output$name), "canada")

tests/testthat/test-create.R

@@ -4,11 +4,7 @@ library(ollamar)
 test_that("create function works with basic input", {
     skip_if_not(test_connection(logical = TRUE), "Ollama server not available")
 
-    expect_error(create("mario"))
-    expect_error(create("mario", modelfile = "abc", path = "abc"))
-    expect_error(create("mario", path = "abc"))
-
-    resp <- create("mario", "FROM llama3\nSYSTEM You are mario from Super Mario Bros.")
+    resp <- create("mario", "deepseek-r1:1.5b")
     expect_s3_class(resp, "httr2_response")
     expect_equal(resp$status_code, 200)
     expect_true(model_avail("mario"))

tests/testthat/test-generate.R

@@ -105,7 +105,7 @@ test_that("structured output", {
     )
 
     msg <- "tell me about canada"
-    resp <- generate("llama3.1", prompt = msg, format = format)
+    resp <- generate("llama3.1:8b", prompt = msg, format = format)
     # response <- httr2::resp_body_json(resp)$response
     structured_output <- resp_process(resp, "structured")
     expect_equal(tolower(structured_output$name), "canada")

tests/testthat/test-ps.R

@@ -8,7 +8,7 @@ test_that("ps list running models endpoint", {
     g1 <- generate('llama3', "tell me a 5 word story")
 
     result <- ps()
-    expect_true(nrow(result) > 1)
+    expect_true(nrow(result) >= 1)
     expect_true(all(c("name", "size", "parameter_size", "quantization_level", "digest", "expires_at") %in% names(result)))
     expect_s3_class(ps("df"), "data.frame")
     expect_s3_class(ps("resp"), "httr2_response")

tests/testthat/test-pull.R

@@ -31,7 +31,7 @@ test_that("pull function works", {
     expect_vector(result$body)
 
     # correct model
-    result <- pull('llama3', stream = TRUE)
+    result <- pull('snowflake-arctic-embed:22m', stream = TRUE)
     # for this endpoint, even when stream = FALSE, the response is chunked)
     expect_true(httr2::resp_headers(result)$`Transfer-Encoding` == "chunked")
     expect_s3_class(result, "httr2_response")
@@ -51,5 +51,7 @@ test_that("pull function works", {
     expect_s3_class(pull('sdafd', stream = TRUE, insecure = TRUE), "httr2_response")
     expect_s3_class(pull('sdafd', stream = TRUE, insecure = FALSE), "httr2_response")
 
+    delete("snowflake-arctic-embed:22m")
+
 })
 

tests/testthat/test-ver.R

@@ -0,0 +1,8 @@
+library(testthat)
+library(ollamar)
+
+test_that("ps list running models endpoint", {
+    skip_if_not(test_connection(logical = TRUE), "Ollama server not available")
+
+    expect_type(ver(), "character")
+})