Allow embeddings requests to be tokens or strings (#417)

* Allow raw tokens to be used as embedding input

* fix linting issues (lines too long)

* add endpoint test for embedding from tokens

* remove redundant comments

* fix comment to match new param name

* change interface to any

* Rename methods and implement convert for base req

* add comments to CreateEmbeddings

* update tests

* shorten line length

* rename parameter

embeddings.go

@@ -113,10 +113,25 @@ type EmbeddingResponse struct {
 	Usage  Usage          `json:"usage"`
 }
 
-// EmbeddingRequest is the input to a Create embeddings request.
+type EmbeddingRequestConverter interface {
+	// Needs to be of type EmbeddingRequestStrings or EmbeddingRequestTokens
+	Convert() EmbeddingRequest
+}
+
 type EmbeddingRequest struct {
+	Input any            `json:"input"`
+	Model EmbeddingModel `json:"model"`
+	User  string         `json:"user"`
+}
+
+func (r EmbeddingRequest) Convert() EmbeddingRequest {
+	return r
+}
+
+// EmbeddingRequestStrings is the input to a create embeddings request with a slice of strings.
+type EmbeddingRequestStrings struct {
 	// Input is a slice of strings for which you want to generate an Embedding vector.
-	// Each input must not exceed 2048 tokens in length.
+	// Each input must not exceed 8192 tokens in length.
 	// OpenAPI suggests replacing newlines (\n) in your input with a single space, as they
 	// have observed inferior results when newlines are present.
 	// E.g.
@@ -129,15 +144,50 @@ type EmbeddingRequest struct {
 	User string `json:"user"`
 }
 
-// CreateEmbeddings returns an EmbeddingResponse which will contain an Embedding for every item in |request.Input|.
+func (r EmbeddingRequestStrings) Convert() EmbeddingRequest {
+	return EmbeddingRequest{
+		Input: r.Input,
+		Model: r.Model,
+		User:  r.User,
+	}
+}
+
+type EmbeddingRequestTokens struct {
+	// Input is a slice of slices of ints ([][]int) for which you want to generate an Embedding vector.
+	// Each input must not exceed 8192 tokens in length.
+	// OpenAPI suggests replacing newlines (\n) in your input with a single space, as they
+	// have observed inferior results when newlines are present.
+	// E.g.
+	//	"The food was delicious and the waiter..."
+	Input [][]int `json:"input"`
+	// ID of the model to use. You can use the List models API to see all of your available models,
+	// or see our Model overview for descriptions of them.
+	Model EmbeddingModel `json:"model"`
+	// A unique identifier representing your end-user, which will help OpenAI to monitor and detect abuse.
+	User string `json:"user"`
+}
+
+func (r EmbeddingRequestTokens) Convert() EmbeddingRequest {
+	return EmbeddingRequest{
+		Input: r.Input,
+		Model: r.Model,
+		User:  r.User,
+	}
+}
+
+// CreateEmbeddings returns an EmbeddingResponse which will contain an Embedding for every item in |body.Input|.
 // https://beta.openai.com/docs/api-reference/embeddings/create
-func (c *Client) CreateEmbeddings(ctx context.Context, request EmbeddingRequest) (resp EmbeddingResponse, err error) {
-	req, err := c.newRequest(ctx, http.MethodPost, c.fullURL("/embeddings", request.Model.String()), withBody(request))
+//
+// Body should be of type EmbeddingRequestStrings for embedding strings or EmbeddingRequestTokens
+// for embedding groups of text already converted to tokens.
+func (c *Client) CreateEmbeddings(ctx context.Context, conv EmbeddingRequestConverter) (res EmbeddingResponse, err error) { //nolint:lll
+	baseReq := conv.Convert()
+	req, err := c.newRequest(ctx, http.MethodPost, c.fullURL("/embeddings", baseReq.Model.String()), withBody(baseReq))
 	if err != nil {
 		return
 	}
 
-	err = c.sendRequest(req, &resp)
+	err = c.sendRequest(req, &res)
 
 	return
 }

embeddings_test.go

@@ -32,6 +32,7 @@ func TestEmbedding(t *testing.T) {
 		BabbageCodeSearchText,
 	}
 	for _, model := range embeddedModels {
+		// test embedding request with strings (simple embedding request)
 		embeddingReq := EmbeddingRequest{
 			Input: []string{
 				"The food was delicious and the waiter",
@@ -46,6 +47,34 @@ func TestEmbedding(t *testing.T) {
 		if !bytes.Contains(marshaled, []byte(`"model":"`+model.String()+`"`)) {
 			t.Fatalf("Expected embedding request to contain model field")
 		}
+
+		// test embedding request with strings
+		embeddingReqStrings := EmbeddingRequestStrings{
+			Input: []string{
+				"The food was delicious and the waiter",
+				"Other examples of embedding request",
+			},
+			Model: model,
+		}
+		marshaled, err = json.Marshal(embeddingReqStrings)
+		checks.NoError(t, err, "Could not marshal embedding request")
+		if !bytes.Contains(marshaled, []byte(`"model":"`+model.String()+`"`)) {
+			t.Fatalf("Expected embedding request to contain model field")
+		}
+
+		// test embedding request with tokens
+		embeddingReqTokens := EmbeddingRequestTokens{
+			Input: [][]int{
+				{464, 2057, 373, 12625, 290, 262, 46612},
+				{6395, 6096, 286, 11525, 12083, 2581},
+			},
+			Model: model,
+		}
+		marshaled, err = json.Marshal(embeddingReqTokens)
+		checks.NoError(t, err, "Could not marshal embedding request")
+		if !bytes.Contains(marshaled, []byte(`"model":"`+model.String()+`"`)) {
+			t.Fatalf("Expected embedding request to contain model field")
+		}
 	}
 }
 
@@ -75,6 +104,15 @@ func TestEmbeddingEndpoint(t *testing.T) {
 			fmt.Fprintln(w, string(resBytes))
 		},
 	)
+	// test create embeddings with strings (simple embedding request)
 	_, err := client.CreateEmbeddings(context.Background(), EmbeddingRequest{})
 	checks.NoError(t, err, "CreateEmbeddings error")
+
+	// test create embeddings with strings
+	_, err = client.CreateEmbeddings(context.Background(), EmbeddingRequestStrings{})
+	checks.NoError(t, err, "CreateEmbeddings strings error")
+
+	// test create embeddings with tokens
+	_, err = client.CreateEmbeddings(context.Background(), EmbeddingRequestTokens{})
+	checks.NoError(t, err, "CreateEmbeddings tokens error")
 }